Research On Mechanical Non-probabilistic Reliability Analysis And Reliability Optimization Design

In mechanical reliability analysis and design, due to the particular characteristics of mechanical parameters, it is difficult to get a great deal of data samples and its margin can be determined easily for those basic data which affect mechanical reliability. It is also difficult to determine distribution function and membership function because it needs more information. The non-probabilistic convex model is adopted to describe uncertainties of mechanical parameters. The response output range can be calculated through state function of machinery whose independent variables is defined in convex set. The index of mechanical reliability can be determined by comparing design safety range with response output range. The non-probabilistic reliability is introduced and it is regarded as a good solution of the above-mentioned problem in the paper. Based on the non-probabilistic convex models, the new methods of non-probabilistic reliability analysis and design are put forward concerning bolt connection, axial pole stability, and interval analysis method of parametric sensitivity in vibration reliability of mechanical components and reliability analysis of axle shaft of machine tool, etc. By means of calculation, the results show the above methods are validity and feasible. Finally, the index of non-probabilistic reliability is introduced in mathematic model of optimization design, and optimizations of axle shaft structure of machine tool and hollow conveying shaft based on non-probabilistic reliability model are presented. The engineering examples show that the presented methods are convenient and effective.