A Study On Experiment Design Method For Structure/Mechanism Reliability

The flexibility of the structure/mechanism joints and the alignment errors significantly affect the structure/mechanism experiment results. The two practical problems are studied in this thesis. The methods to solve the problems are given, parameters design of the simulation joint and simulation design of the alignment errors are achieved. In order to compare the difference between theoretical calculation method of structure failure probability and practical experiment result, a reliability experiment is done, and a plan of the structure system reliability experiment is discussed. The following work has been done in this paper:The load characteristic and design calculation method of spring are systematically studied. The method of using assembled springs to simulate flexible joint is presented, and then the structural configuration of simulation joint is given. The method above-developed can be used to calculate the parameters of the simulation joint on the given load and displacement conditions. An example of parameter design of a simulation joint is given.The effect of the alignment error for structure/mechanism reliability is analyzed. Based on the randomicity of the alignment errors, the method called "pseudo random number" is presented to simulate pseudo alignment errors. The theory of producing pseudo random number has been studied systematically. The method above-developed can be used to simulate pseudo alignment errors, and an example illustrates the method.For a five bars truss structure, the introduction of the basic method of structural failure probability analysis is explained. Based on O.Ditlevsen's narrow bounds' algorithm and the numerical integration method in "M" space, a computer program to calculate failure probability of the five poles truss structure is given.Based on the statistic values of the bars' strength fracture experiment, the failure probability theoretical value of five bars truss structure is obtained. With the aid of the theory of the incremental load method [ICM], the failure probability experiment plan of the five bars truss structure is discussed. Some problems of using the ICM in thefive bars truss structure reliability experiment are analyzed, and some author's opinions are presented.