Fatigue Reliability Analysis of Motor Hanger for High-Speed Train Based on Bayesian Updating and Subset Simulation

Li, Yonghua; Zhi, Pengpeng; Zhang, Yue; Chen, Bingzhi; Wang, Yuedong

doi:https://doi.org/10.1155/2020/3012471

Advances in Materials Science and Engineering

On this page

Abstract Introduction Conclusions Data Availability Conflicts of Interest Acknowledgments References Copyright Related Articles

Special Issue

Advances in Structural Integrity and Lifetime Prediction of Power & Energy Systems

View this Special Issue

Research Article | Open Access

Volume 2020 | Article ID 3012471 | https://doi.org/10.1155/2020/3012471

Fatigue Reliability Analysis of Motor Hanger for High-Speed Train Based on Bayesian Updating and Subset Simulation

Yonghua Li,¹Pengpeng Zhi,²Yue Zhang,¹Bingzhi Chen,¹and Yuedong Wang¹

Guest Editor: Shun-Peng Zhu

Received31 Jan 2020

Accepted27 Feb 2020

Published27 Mar 2020

Abstract

In order to more accurately analyze the fatigue reliability of motor hanger for high-speed train and reduce the influence of uncertain factors, a Bayesian statistical method is introduced to propose a novel fatigue reliability analysis method based on Bayesian updating and subset simulation. First, considering the influence of various uncertain parameters on the first principal stress (FPS) of motor hanger, the ANSYS parametric design language (APDL) is used to establish the parametric model. The D-optimal design of experiment is carried out to calculate the FPS of the motor hanger. Second, the experimental data is fitted by the least square method to establish a polynomial response surface function which characterizes the FPS of the motor hanger, and analysis of variance (ANOVA) is carried out. On this basis, the variation trend of the FPS under parameter fluctuation is calculated, and its probability distribution characteristics are obtained. Based on the MATLAB platform, the Bayesian updating method is adopted to correct the probability and statistical characteristics of the FPS to improve the accuracy of prediction. Finally, the subset simulation (SS) method is used to calculate the fatigue failure probability of the motor hanger. The research results show that the proposed method helps to improve the accuracy and efficiency of fatigue reliability analysis.

1. Introduction

In the design process of mechanical products, the analysis of structural static strength and fatigue strength is the key to ensure that the product meets the standard requirements. However, with the continuous improvement of the operation speed of high-speed trains, in order to ensure the safety and reliability of operation, the design of its key components must meet the requirements of antifatigue design and reliability [1]. Currently, the evaluation of the fatigue performance of key components for high-speed trains is mainly based on fatigue strength and cumulative fatigue damage. The evaluation index can be used to perform fatigue strength analysis and fatigue life prediction on the structure quickly and easily and is easy to compare with the experimental results. But, the disadvantages of this method are that the results of evaluation and analysis are too ideal to consider the design tolerance and manufacturing error, and the comparison with the experimental results is too single and not universal, which is not enough to reflect the fatigue reliability of all products. In response to the above problems, some scholars have combined the reliability theory and fatigue analysis theory to study the fatigue reliability of mechanical products and proposed some fatigue reliability analysis methods [2–12]. Chen et al. [13] proposed a fatigue reliability analysis method that considers the uncertainty of parameters, draws a fatigue limit diagram without safety factor, and gives a more reasonable fatigue reliability analysis result, which is helpful for lightweight design of products. Li et al. [14] established a robust optimization model based on Six Sigma to analyze the fatigue reliability of the pantograph collector head support. This method not only considers the influence of uncertain factors, but also improves the robustness of fatigue life prediction. Bayraktar et al. [15] considering the number of kilometers and load cycles, statistically evaluated the actual life value. And the theoretical and practical Wohler diagrams S-N is plotted to evaluate the fatigue reliability of the axle. Zhu et al. [16] presented a framework for fatigue reliability assessments and service life prediction based on the estimation of the evolution and probabilistic distribution of fatigue damage over time, and the effectiveness of the method is verified by taking the axle of train as an example. An accelerated life test (ALT) method is proposed by Lu et al. which is first employed to predict the fatigue life of a full-scale bogie frame [17]. This method provides a reference for reliability evaluation of large structures. Zuo et al. [18] studied the fatigue damage of the bogie frame and used the Bayesian updating method to analyze the fatigue reliability of the welded bogie frame. Hu et al. [19] combined finite element method with experiment to evaluate the fatigue reliability of bolster under the application of the design passenger number spectra based on stress-life interference model. The above studies have played an important role in fostering applications of reliability theory for fatigue problems in the structural safety field. For the design process, how to accurately evaluate the fatigue reliability of the structure without experimentation is the premise of predicting whether the design meets the requirements of the standard. Generally, reliability methods based on simulation or approximation have three types: (1) sampling-based methods, (2) moment matching methods, and (3) MPP-based methods. Sampling-based methods are easy to apply and can provide accurate probability estimations with sufficient simulations [20]. Thus, a Bayesian updating combined with subset simulation method is proposed in this article to address the problem.

The contributions of this study are as follows: (1) a response surface function is presented to characterize the functional relationship between random variables and the FPS of the structure; (2) the probabilistic and statistical parameters of the FPS are updated by using Bayesian theory; (3) the limit state equation of fatigue reliability analysis is established; (4) the fatigue failure probability of the motor hanger is calculated by the subset simulation method.

The reminder of the paper is organized as follows. Section 2 reviews Bayesian updating and subset simulation method; we used both methods in fatigue reliability analysis. Section 3 provides details on the proposed approach. Section 4 uses the fatigue reliability analysis of the motor hanger as an example to illustrate its effectiveness and practicality. Finally, Section 5 summarizes and concludes.

2. Bayesian Updating Theory and Subset Simulation Method

2.1. Bayesian Updating Theory

In engineering practice, it is often necessary to make decisions and processes on the data that has already been obtained. These data may include observational data, experimental data, and simulation data. When acquiring new data, it is necessary to update the acquired data to ensure the reliability of the data. In view of the above problems, this paper uses Bayesian method to update and process the uncertainty and difference of simulation data [21, 22].

The basic principle of Bayesian theory can be roughly expressed as using prior distribution and sample information to solve its posterior distribution; that is, the posterior distribution is obtained based on prior information and experimental data [23]. The specific steps for solving it are as follows:(1)Determine the parameters that need to be solved and set them as random variables(2)According to the sample data, the corresponding likelihood function and prior distribution function are constructed(3)The posterior distribution function is obtained by Bayes equation, and the posterior distribution function is used to solve various reliability indexes

Because Bayes method makes full use of prior information and experimental data, the posterior distribution function is more consistent with the probability and statistical characteristics of random variables. The traditional Bayesian equation is given bywhere is the posterior distribution function; is the prior distribution function; and is the edge distribution function, and the two types are and , respectively.

The distribution types of design parameters of key components for high-speed railway are mainly lognormal distribution and normal distribution, and the two probability distributions can be converted to each other. However, the FPS used for fatigue reliability analysis mainly obeys logarithmic distribution. Therefore, the prior distribution of lognormal distribution should be transformed into the prior distribution of normal distribution. The transformation equation can be expressed aswhere and represent the mean and standard deviation of prior lognormal distribution, respectively, and and represent the mean and standard deviation of prior normal distribution, respectively.

According to Bayesian theory [24], the posterior distribution function of normal distribution can be obtained from equations (2) and (3), which is written aswhere and represent the mean and standard deviation of posterior normal distribution, respectively; and represent the mean and standard deviation of the normal distribution actually obtained from the evaluated parameter, respectively; represents the number of statistical samples; and represents the standard deviation caused by uncertainties.

Combining equations (2)–(5), the statistical characteristic parameters expression of the evaluated parameters is obtained by Bayesian updating, which is given bywhere and represent the mean and standard deviation of lognormal distribution after Bayesian updating, respectively.

2.2. Subset Simulation Method

The subset simulation method was first proposed by Au and Beck [25], and it is a method that can efficiently deal with high-dimensional and small probability problems. The method can represent the failure probability with a small value as the product of a series of large condition failure probabilities; that is, the simulation process of the small failure probability event is transformed into a series of large event simulation processes. If the intermediate failure event satisfies the nested relationship (), the target failure probability can be rewritten aswhere is the target failure event, , and is the critical value of the structural response.

In the reliability analysis, subset simulation starts with direct Monte Carlo in the first step. The probability is estimated aswhere is an independent and identically distributed sample generated by simulation of joint probability density function with input random vectors; N is the number of sample points per layer; and is indicator function.

Similarly, the other conditional probabilities are calculated by the idea of equation (9). In the modified Metropolis algorithm, a group of one-dimensional proposal PDFs are used, instead of an n-dimensional proposal PDF which is used in Metropolis algorithm. Thus, the acceptance ratio of individual sample can remain nonvanishing in spite of the increasing of dimension [26, 27]. Thus, in order to efficiently generate conditional samples, this paper implements MCMC simulation based on improved Metropolis–Hastings algorithm.

3. The Analysis Method of Fatigue Reliability

3.1. Fatigue Life Prediction

Fatigue, as the main reason for failure of engineering structures and components, is the main factor affecting the safety of high-speed train operation. Therefore, more and more attention has been paid to antifatigue design in the design process. At present, the prediction of structural fatigue life is mainly based on the S-N curve or the P-S-N curve. The S-N curve represents the stress-life curve, and the P-S-N curve is the stress-life curve of the material under different survival conditions. The most commonly used form of S-N curve describing materials is the power function form, as follows:

Equation (11) can be rewritten aswhere m and C are parameters related to material, stress ratio, and loading mode; S represents stress; and N represents lifetime.

According to the design requirements, the material suitable for the motor hanger of high-speed railway trains is S355J2G. However, the mechanical properties of S355J2G are difficult to obtain. For this reason, the material parameters of Q345 similar to its mechanical properties are selected for fatigue reliability analysis [28]. Table 1 gives the main parameters of the material P-S-N curve, and the P-S-N curve of the material is plotted according to its parameters, as detailed in Figure 1.

3.2. Limit State Equation for Fatigue Reliability Analysis

In this paper, the fatigue lifetime model is selected for fatigue reliability analysis. The model uses fatigue lifetime as the parameter for reliability calculation. The limit state equation can be expressed aswhere N is the calculated fatigue lifetime and N₀ is the design lifetime; when Z > 0, the structure is in a safe state; when Z = 0, the structure is in a limit state; when Z < 0, the structure is in a failure state.

From the definition of reliability, the fatigue reliability of the structure is the probability of the function Z > 0, and the expression of the reliability is

The failure probability of a structure can be expressed as

4. Fatigue Reliability Analysis of Motor Hanger for High-Speed Train

Fatigue reliability analysis is an analytical method to improve product quality by combining fatigue lifetime analysis and reliability design. It is essential to ensure that the structural design meets the standard requirements. As the load-bearing structure between the motor and the bogie frame, whether the performance and service time of the structure can meet the requirements is directly related to the normal operation of the train. To this end, this paper takes the motor hanger as the research object and uses Bayesian updating and subset simulation to explore its fatigue reliability, which provides a new idea and method for the fatigue life prediction in the design stage. The specific process of fatigue reliability analysis is provided in Figure 2.

4.1. Finite Element Analysis of Motor Hanger

The premise of fatigue reliability analysis for motor hanger is to calculate its FPS under fatigue conditions. The research object of this paper is the motor hanger of high-speed train, which is bolted to the bogie frame to fix the motor. The motor hanger is mainly plate welded, and the partial opening is used to reduce the quality. Figure 3 shows the specific installation position and structure type. The numerical analysis by the finite element method is performed to obtain the FPS of the motor hanger. It is modeled by shell and solid mode, given in Figure 4. The motor hanger is modeled using beam188, shell181, solid 185, rigid, and rbe3 elements. It is meshed to have 50,879 elements, including 21,766 triangular and quadrilateral shell elements and 28,973 tetrahedral and hexahedral solid elements.

In the finite element model, the reasonable application of load cases is the key to ensure the accuracy of the analysis results. Generally, the boundary conditions of the test components under normal operating conditions should be consistent with those under test assembly conditions. The elastic connector on the bogie can be replaced by a rigid member, but the direction and magnitude of the load acting on the test piece cannot be changed. Figure 5 displays the load direction of motor hanger, which is positive upward. The loads are calculated according to EN13749 standard and analyzed by finite element method. Figure 6 plots the result of the finite element analysis. It can be clearly seen from the figure that the maximum FPS is 119.23 MPa, which occurs on the motor boom.

4.2. Establishment of Response Surface Function

The purpose of establishing the response surface function is to obtain the functional relationship between the random variables and the FPS and then to observe the influence of the uncertainties on it. As one of the surrogate models, the response surface can more accurately represent the functional relationship between input and output. Compared to other surrogate models, the response surface belongs to the display function that can be more easily used for reliability calculations [29]. Due to the poor approximation ability of the linear response surface, the quadratic response surface with cross terms is adopted. The basic equation is given bywhere is the number of uncertain variables, is the constant, and , , and are the polynomial coefficients, respectively.

According to the stress plot in Figure 6, the random variables which have great influence on the stress plot are selected as random variables. Considering the uncertainty of design parameters, the upper and lower limits of size parameters are set according to tolerance, and the upper and lower limits of load are obtained according to test conditions. Table 2 displays the specific parameters for random variables.

The premise of establishing response surface function is to do the design of experiment for random variables. According to the design of experiment scheme, the finite element model is invoked to solve the problem, and the corresponding response of each scheme is obtained. In this paper, the D-optimal design of experiment method is used to carry out the experimental design; see Table 3.

According to (16), the experimental data in Table 3 is fitted by least squares method to obtain a response surface function, which characterizes the function relationship between the random variable and the FPS.

The fitting accuracy of the response surface function is the basis and premise to ensure the accuracy of the subsequent reliability analysis. According to the fitting result , indicating that the fitting accuracy is high. To further verify the accuracy of the response surface function, the ANOVA is adopted to analyze the function in this paper. Some results are provided in Table 4.

In Table 4, the F and P values are the main indicators used to evaluate the significance of the response surface function and its random variables. The larger the F value, the more significant it is; on the contrary, the smaller the P value, the more significant it is. As can be seen from Table 4, the function generally presents a very significant level, , indicating that the fitting accuracy of response surface is high. Among the four random variables, random variable TH1 is the most significant, which indicates that the small fluctuation of the random variable TH1 has a greater impact on the FPS. The other three random variables did not exhibit a very significant level, only that their small fluctuations had little effect on the FPS. Because of space, the interaction among random variables cannot be given. Through analysis, it is found that the interaction among random variables in response surface function is relatively low, which indicates that the interaction between random variables has less influence on response. By comparing the results of the first and second significance of random variables, it can be seen that the second significance of TH1 decreases and the second significance of TH2, F₁, and F₂ increases, indicating that the selection of random variables is reasonable. In summary, the response surface function can be used for fatigue reliability analysis of subsequent motor hanger.

4.3. Determination of the FPS Distribution Type and Bayesian Updating

For equation (13), the uncertainty of the calculated lifetime N results in three different states of the structure. As the main factor affecting the calculated lifetime N, accurate probability distribution of the FPS is the key to improve the accuracy of reliability analysis. For this reason, the FPS under random parameter fluctuation is obtained by using Monte Carlo method for 10000 calculations based on equation (17). The FPS is fitted by MATLAB 2015b to determine its probability distribution characteristics, as shown in Figure 7. From Figure 7, it can be seen that the FPS obeys lognormal distribution, and the fitting curve has high accuracy, . In order to accurately estimate the fatigue reliability of the structure, simulation data are used as real measurement data for Bayesian updating because there is no experimental data in the design process. The prior distribution data of the FPS are obtained from the experimental data of similar products. The data of prior distribution obey the lognormal distribution of , and the real measured data obey the lognormal distribution of . Two kinds of distribution data are substituted into equations (2)–(7) to obtain the probability and statistical parameters of the FPS after Bayesian updating. Figure 8 plots the probability distribution of the FPS before and after Bayesian updating. It can be seen from Figure 8 that the mean of FPS value after the Bayesian updating is slightly smaller than that of the FPS obtained by the actual engineering, and the corresponding variance is also reduced. This proves that the Bayesian updating method not only reduces the uncertainty of FPS, but also reduces the distortion and uncertainty of the detected data. In the design sense, Bayesian updating reduces the uncertainty of the simulation results.

4.4. Fatigue Reliability Analysis Based on Subset Simulation

The FPS of the motor hanger after Bayesian updating obeys the lognormal distribution of . The symmetrical cyclic load spectrum specified in the EN13749 standard is selected, and the stress ratio is R = -1. The P-S-N curves with survival rates of 95% and 99% are selected, respectively, and the design lifetime is 10 million cycles. According to equations (11)–(13), the fatigue failure probability of the motor hanger is calculated by SS. Figures 9 and 10 display the CDF curves for fatigue failure probabilities at different survival rates, respectively. The estimated fatigue failure probabilities and the number of required samples are listed in Table 5.

It can be seen from Figures 9 and 10 that the overall fatigue failure probability with survival rate of 99% is higher than the survival rate of 95%, indicating that the selection of the material P-S-N curve has a direct impact on the fatigue reliability analysis results of the structure, and should be paid attention to during the design process. Then, Figures 9 and 10 plot the CDF for = 0.15, 0.20, 0.25, and 0.30. The purpose of this comparison is to see how SS behaves differently by adopting different values of for the fatigue reliability analysis of motor hanger; is the conditional failure probability of subset simulation. In Figure 9, the CDF plot with = 0.3 significantly deviates from the rest. However, the CDF plot with = 0.1 significantly deviates from the rest. It can be seen that the selection of is very important for estimating the fatigue failure probability. Although Zuev et al. [30] proposed the values of selected from the interval [0.1, 0.3] would produce similar efficiency with similar accuracy, the accuracy is still different for different subjects. For the fatigue reliability analysis of the motor hanger, when the survival rate is 95%, the failure probability obtained by Monte Carlo simulation is 0.0015; = 0.10, 0.15, 0.20, and 0.25 are suitable for reliability analysis, while = 0.30 is not proper. Similarly, when the survival rate is 99%, the failure probability obtained by Monte Carlo simulation is 0.1880; = 0.15, 0.20, and 0.25 are suitable for reliability analysis, while = 0.1 and 0.30 are not proper. For the calculation of Monte Carlo failure probability, 10 000 sample points are used, while the SS method is only 4%–7% of Monte Carlo, which greatly improves the calculation efficiency.

5. Conclusions

This study presents a novel fatigue reliability analysis method based on Bayesian updating and subset simulation to improve the accuracy of structural fatigue performance evaluation at the design stage. In order to reduce the uncertainty of fatigue lifetime calculated by simulation analysis, the Bayesian method is used to update the probability and statistical parameters of the FPS for the motor hanger. Compared with the original data, the mean and variance of the FPS after the update are reduced, which improves the robustness of the data. Meanwhile, the SS method is adopted to calculate the fatigue failure probability under different survival rates and different conditions failure probabilities, and the CDF curve of fatigue failure probability is given. The fatigue failure probability of motor hanger under P-S-N curve with 95% survival rate is = 0.1 = 0.005, = 0.15 = 0.0055, = 0.20 = 0.005, = 0.25 = 0.005. The fatigue failure probability of motor hanger under P-S-N curve with 99% survival rate is = 0.15 = 0.15, = 0.20, = 0.1950, = 0.25 = 0.1950. This not only proves the influence of the P-S-N curve with different survival rates on the fatigue failure probability, but also verifies that the reasonable selection of conditional failure probability is helpful to improve the accuracy of the SS. In a word, the proposed method helps the design stage to more accurately assess whether the fatigue lifetime of the structure meets the standard requirements. At the same time, it is also applicable to the fatigue reliability analysis of other large complex structures.

Data Availability

The data used to support the findings of this study are included within the article.

Conflicts of Interest

The authors declare that there are no conflicts of interest regarding the publication of this article.

Acknowledgments

The authors would like to acknowledge the National Natural Science Foundation of China under Contract no. 51875073, the Scientific Research Project of Education Department of Liaoning Province under Contract no. JDL2019005, the Dalian Science and Technology Innovation Fund Plan under Contract no. 2019J11CY017, and the Innovation Team Support Plan of Liaoning Higher Education under Contract no. LT2016010.

References

P. Zhi, Y. Li, and B. Chen, “Structural strength analysis of bogie frames considering parameter uncertainty,” China Mechanical Engineering, vol. 30, no. 1, pp. 22–29, 2019.
View at: Google Scholar
S. Zhu, B. Keshtegar, N. Trung, Z. Yaseen, and D. Bui, “Reliability-based structural design optimization: hybridized conjugate mean value approach,” Engineering with Computers, 2019, In press.
View at: Publisher Site | Google Scholar
B. Keshtegar and S.-P. Zhu, “Three-term conjugate approach for structural reliability analysis,” Applied Mathematical Modelling, vol. 76, pp. 428–442, 2019.
View at: Publisher Site | Google Scholar
D. Meng, S. Yang, Y. Zhang, and S. P. Zhu, “Structural reliability analysis and uncertainties-based collaborative design and optimization of turbine blades using surrogate model,” Fatigue & Fracture of Engineering Materials & Structures, vol. 42, no. 6, pp. 1219–1227, 2019.
View at: Publisher Site | Google Scholar
S.-P. Zhu, Q. Liu, W. Peng, and X.-C. Zhang, “Computational-experimental approaches for fatigue reliability assessment of turbine bladed disks,” International Journal of Mechanical Sciences, vol. 142-143, pp. 502–517, 2018.
View at: Publisher Site | Google Scholar
R. F. Sanches, A. M. P. de Jesus, J. A. F. O. Correia, A. L. L. da Silva, and A. A. Fernandes, “A probabilistic fatigue approach for riveted joints using Monte Carlo simulation,” Journal of Constructional Steel Research, vol. 110, pp. 149–162, 2015.
View at: Publisher Site | Google Scholar
R. Liu, P. Chen, X. Zhang, and S. Zhu, “Non-shock ignition probability of octahydro-1,3,5,7-tetranitro-tetrazocine-based polymer bonded explosives based on microcrack stochastic distribution,” Propellants, Explosives, Pyrotechnics, 2019, In press.
View at: Publisher Site | Google Scholar
S. P. Zhu, Q. Liu, J. Zhou, and Z. Y. Yu, “Fatigue reliability assessment of turbine discs under multi-source uncertainties,” Fatigue & Fracture of Engineering Materials & Structures, vol. 41, no. 6, pp. 1291–1305, 2018.
View at: Publisher Site | Google Scholar
M. Miguel, M. Abílio, J. De, A. José, and F. Alfonso, “A methodology for probabilistic prediction of fatigue crack initiation taking into account the scale effect,” Engineering Fracture Mechanics, vol. 185, pp. 101–113, 2017.
View at: Google Scholar
S. Zhu, Q. Liu, Q. Lei, and Q. Wang, “Probabilistic fatigue life prediction and reliability assessment of a high pressure turbine disc considering load variations,” International Journal of Damage Mechanics, vol. 27, no. 10, pp. 1569–1588, 2018.
View at: Google Scholar
D. Meng, Y. Li, S.-P. Zhu, G. Lv, J. Correia, and A. de Jesus, “An enhanced reliability index method and its application in reliability-based collaborative design and optimization,” Mathematical Problems in Engineering, vol. 2019, Article ID 4536906, 10 pages, 2019.
View at: Publisher Site | Google Scholar
Y. Ai, S. P. Zhu, D. Liao et al., “Probabilistic modeling of fatigue life distribution and size effect of components with random defects,” International Journal of Fatigue, vol. 126, pp. 165–173, 2019.
View at: Publisher Site | Google Scholar
B. Chen, P. Zhi, and Y. Li, “Fatigue strength analysis of bogie frame in consideration of parameter uncertainty,” Frattura Ed Integrità Strutturale, vol. 13, no. 48, pp. 385–399, 2019.
View at: Publisher Site | Google Scholar
Y. Li, M. Hu, and F. Wang, “Fatigue life analysis based on six sigma robust optimization for pantograph collector head support,” Advances in Mechanical Engineering, vol. 8, no. 11, pp. 1–9, 2016.
View at: Publisher Site | Google Scholar
M. Bayraktar, N. Tahrali, and R. Guclu, “Reliability and fatigue life evaluation of railway axles,” Journal of Mechanical Science and Technology, vol. 24, no. 3, pp. 671–679, 2010.
View at: Publisher Site | Google Scholar
S.-P. Zhu, H.-Z. Huang, Y. Li, Y. Liu, and Y. Yang, “Probabilistic modeling of damage accumulation for time-dependent fatigue reliability analysis of railway axle steels,” Proceedings of the Institution of Mechanical Engineers, Part F: Journal of Rail and Rapid Transit, vol. 229, no. 1, pp. 23–33, 2015.
View at: Publisher Site | Google Scholar
Y. Lu, H. Zheng, J. Zeng, T. Chen, and P. Wu, “Fatigue life reliability evaluation in a high-speed train bogie frame using accelerated life and numerical test,” Reliability Engineering & System Safety, vol. 188, pp. 221–232, 2019.
View at: Publisher Site | Google Scholar
F. Zuo, Y. Li, and H. Huang, “Reliability analysis for fatigue damage of railway welded bogies using Bayesian update based inspection,” Smart Structures and Systems, vol. 22, no. 2, pp. 193–200, 2018.
View at: Google Scholar
J. Hu, L. Xie, W. Sun, X. Li, X. Lv, and W. Meng, “Structural safety and fatigue reliability assessments of a metro-train bolster,” Advances in Mechanical Engineering, vol. 11, no. 1, pp. 1–11, 2019.
View at: Publisher Site | Google Scholar
D. Meng, H.-Z. Huang, Z. Wang, N.-C. Xiao, and X.-L. Zhang, “Mean-value first-order saddlepoint approximation based collaborative optimization for multidisciplinary problems under aleatory uncertainty,” Journal of Mechanical Science and Technology, vol. 28, no. 10, pp. 3925–3935, 2014.
View at: Publisher Site | Google Scholar
S.-P. Zhu, H.-Z. Huang, W. Peng, H.-K. Wang, and S. Mahadevan, “Probabilistic physics of failure-based framework for fatigue life prediction of aircraft gas turbine discs under uncertainty,” Reliability Engineering & System Safety, vol. 146, pp. 1–12, 2016.
View at: Publisher Site | Google Scholar
S.-P. Zhu, H.-Z. Huang, R. Smith, V. Ontiveros, L.-P. He, and M. Modarres, “Bayesian framework for probabilistic low cycle fatigue life prediction and uncertainty modeling of aircraft turbine disk alloys,” Probabilistic Engineering Mechanics, vol. 34, pp. 114–122, 2013.
View at: Publisher Site | Google Scholar
K. C. Siju and M. Kumar, “Bayesian Estimation of reliability using time-to-failure distribution of parametric degradation models,” Journal of Statistical Computation and Simulation, vol. 88, no. 9, pp. 1717–1748, 2018.
View at: Publisher Site | Google Scholar
D. M. Frangopol, “Probability concepts in engineering: emphasis on applications to civil and environmental engineering,” Structure and Infrastructure Engineering, vol. 4, no. 5, pp. 413-414, 2008.
View at: Publisher Site | Google Scholar
S.-K. Au and J. L. Beck, “Estimation of small failure probabilities in high dimensions by subset simulation,” Probabilistic Engineering Mechanics, vol. 16, no. 4, pp. 263–277, 2001.
View at: Publisher Site | Google Scholar
De. Meng, Y. Li, H. Huang, Z. Wang, and Y. Liu, “Reliability-based multidisciplinary design optimization using subset simulation analysis and its application in the hydraulic transmission mechanism design,” Journal of Mechanical Design, vol. 137, no. 5, pp. 051402–51411, 2015.
View at: Publisher Site | Google Scholar
D. Meng, Y. Li, S. Zhu, Z. Hu, T. Xie, and Z. Fan, “Collaborative maritime design using sequential optimisation and reliability assessment,” Proceedings of the Institution of Civil Engineers-Maritime Engineering, 2020, In press.
View at: Publisher Site | Google Scholar
P. Zhi, Y. Xu, and B. Chen, “Time-dependent reliability analysis of the motor hanger for EMU based on stochastic process,” International Journal of Structural Integrity, 2020, In press.
View at: Publisher Site | Google Scholar
D. Angın and A. Tiryaki, “Application of response surface methodology and artificial neural network on pyrolysis of safflower seed press cake,” Energy Sources, Part A: Recovery, Utilization, and Environmental Effects, vol. 38, no. 8, pp. 1055–1061, 2016.
View at: Google Scholar
K. M. Zuev, J. L. Beck, S.-K. Au, and L. S. Katafygiotis, “Bayesian post-processor and other enhancements of subset simulation for estimating failure probabilities in high dimensions,” Computers & Structures, vol. 92-93, pp. 283–296, 2012.
View at: Publisher Site | Google Scholar

Copyright

Copyright © 2020 Yonghua Li et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

PDF Download Citation

Download other formats

Order printed copies

Views

501

Downloads

902

Citations