Reliability Analysis Of Structural Systems Consider Static Strength, Fatigue And Stability Simultaneously

For piezoelectric structure system, there are a lot of uncertainty factors, for example, the loads which act on the structure are indeterminate; material data and dimension data provided may be not consistent with that of real structure; Therefore, it is necessary for piezoelectric structure to make stochastic analysis and evaluate its reliability. Static strength reliability, stable reliability, dynamic reliability and reliability based on Neural Networks are analyzed by using the theory of reliability for truss structure with piezoelectric patches affixed to its surface in this paper. The main research works of the paper are as following1) For the truss structure comprised of piezoelectric active-member with piezoelectric patches affixed to its surface, the finite element model is founded, which includes static force models, stabile models and dynamical models. The responses of internal force under Mech-Electric coupling are calculated.2) The failure mechanism of the piezoelectric member is discussed. In the process of statics and dynamic reliability analysis, the failure criterion of the piezoelectric member is proposed, safety margin functions of failure modes which are caused by yield failure of piezoelectric elementary body, fracture damage of piezoelectric patches, clue layer destroying and electrical breakdown are established separately, In the process of stabile reliability analysis, safety margin functions which are composed of load coefficient are established and stable reliability analysis method of piezoelectric truss is proposed in different condition.3) By considering the yield strength of piezoelectric elementary body, area of section, fracture strength and breakdown electric field intensity of the piezoelectric member, loads as stochastic variables, the formula of partial derivative of the safety margin functions is studied, the method of reliability of safe margin is given by adopting Taylor expansion SFEM; Then adopting basic theory of reliability, combining the mechanical character of piezoelectric truss structure, the reliability analysis of piezoelectric truss structure is discussed. At last, the reliability index of piezoelectric truss structure system is got by PNET.4) By means of Neural Networks which has strong ability of function mapping and using orthogonal table as sample set, combining with Monte Carlo stochastic sampling, the function values of limit state are obtained. Reliability analysis of piezoelectric truss structure based on Neural Network is discussed also.5) For program compiling, the static force, stabile and dynamical reliability effect of Mech-Electric coupling and external voltage on piezoelectric truss structure are analyzed. The results indicates the reliability of piezoelectric truss structure is better than common structure, and the effect of optimum distribution of piezoelectric member to the reliability of structural system is analyzed.