| In the statistically independent normal distribution space,there are many methods for calculating the structural reliability index,such as the center point method,the improved first order second moment method,the gradient optimization method,the Monte Carlo method,and the JC method.This article focuses on these five commonly used traditional algorithms for analysis.When calculating structural reliability indicators,the relevant literature mentions that when using the Mean First Order Second Moment Method,serious errors may occur for the different types of structural performance functions and their functional forms.Improvement of a second-order moment method overcomes the difficulties of the former.Then,it is rarely detailed that,whether or not it will be affected by some unknown factors in the calculation using the remaining methods.For the above problems,and based on the needs of basic teaching and academic research,this paper comparatively studied the calculation accuracy of the above five algorithms.Through an example analysis,the effects of the five algorithms on the reliability index calculation results under the conditions of different forms and types of structural performance functions and in the changing reliable space are compared in detail,which is using the Fortran language for programming calculations.The results show that when the structural reliability index is calculated in a statistically independent normal distribution space,the calculation result of the Mean First Order Second Moment Method is affected most by the type and form of the structural performance function.In addition,the results of the Gradient Optimization Method and the JC method based on Orthogonal Transformation are not affected by the form and type of the structural performance function,but in the case of reliability less than 0.5,their calculation results will have a large error;However,the results of the Improved First Order Second Moment Method Based on Rackwitz Iterations are not affected by the types,forms,and low reliability of the function,and their convergence and calculation accuracy are good.Therefore,it is proposed to use the JC method to avoid the use of "Orthogonal transformation" and avoid the use of "Gradient Optimization method" when solving structural reliability.In order to evaluate the reliability of a excavator's boom structure during the continuous operation of the load,the Boom of a model domestic excavator is taken as the object of study.After the structure simplified,ANSYS finite element analysis software is used to construct the calculation model.Through the load analysis,static stress analysis is carried out to extract the maximum stress of the structure.Taking into account the impact of different environments on the excavator operation,the stress samples of the operation are analyzed by the W test method.A dynamic reliability solution method that is considering the degradation of mechanical strength was established by using the theory of the "Residual Strength" and "Reliability theory considering the influence of the number of loads" to calculate the reliability at the dangerous point.Through a MATLAB's programming calculation,the dynamic reliability of the continuous operation number within 8 hours of the boom during different use periods is given.The dynamic reliability result of the boom considering the residual strength calculated by this method is more meaningful than the "single reliability value" and "fatigue life value" of the excavator boom that was calculated by the fatigue reliability method.In this paper,the relationship between the system reliability model under different loads and the multivariate nonlinear function in the traditional algorithm are studied,and a diversified system reliability model is established,which is based on the comparison of the traditional model of reliability integral calculation and the characteristics of the system layer R-S model.Based on the multivariate nonlinear function in the traditional algorithm,the idea of the joint condition reliability is proposed according to the physical meaning of the stress function,and the probability density function including the diversity information of the components and the environment is established,which is the new condition reliability.Relying on the Stress-Strength Function model,a multi-dimensional reliability model is established by considering not only the influence of various components but also the influence of different loads.Finally,the Monte Carlo method and the traditional PNET method are used to verify the reliability model,and the result proves the correctness of the model after this diversification.According to some basic characteristics of a megawatt steel wind turbine tower such as the elevation dimension,its mechanical model is established by using ANSYS software.The structure is a tubular shape with a height of 44.075 meters,and its cross-section and shell thickness are variable along its height.In order to examine the dynamic reliability of the structure over its expected life,a corresponding finite element model has been established.According to the analysis of the statistical results of the 18 cases of wind load in 20 years,the Poisson random process intensity parameter ?(t)of the wind load is calculated based on the Poisson stochastic process.Using an existing "dynamic reliability model without considering the strength degradation",the extreme load of wind and the gravity of the tower are taken as the main load combinations of the structure,and the dynamic reliability of the structure is calculated through a MATLAB programming under the assumption of two distributions of the structural strength.Finally,the dynamic reliability of the megawatt wind turbine steel tower is obtained during the estimated 20-year service period,which can be expected to provide a data reference for the manufacture of this type of equipment and the management and maintenance of the equipment during the service period.At the same time,a preliminary exploration of algorithmic application and preliminary analysis of algorithm performance are made for the recent reliability analysis of large-scale structures.The research results and conclusions of this paper have certain significance for the selection and application of structural reliability algorithms. |
PDF Full Text Request