Optimal Design Of Crane Metal Structure Based On Non-probabilistic Reliability

Crane is used to reduce physical labor, improve working efficiency and realize safe production as an auxiliary equipment and widely used in machinery manufacturing, ports, construction, transportation and other areas of production. Crane metal structure is an important part of lifting equipment, which is used to inherit and pass work load and natural load. So the design of crane metal structure has a great effect on the mechanical performance.Traditional crane structure optimization design is based on the deterministic, structure parameters have certain uncertainty in practical engineering, which may cause the crane does not meet the design requirements. probabilistic reliability theory can deal with the uncertainty, but it needs enough dates to describe the probability distribution of the uncertain parameters, sometimes it is difficult to obtain. Non-probabilistic reliability convex theory describes the uncertain parameters with interval model, it only get the boundary of the parameters and has simple structure and strong applicability. Based on reliability analysis of interval model, the limitation state function for crane structure is difficult to obtain directly. A structural optimization method based on response surface model and non-probabilistic reliability analysis is presented and the crane structure is optimized for its minimum volume in this paper.Firstly, based on the basic theory of the response surface, quadratic polynomial is adopted to replace approximately the limitation state function of the structure, then using ellipsoid model and interval model to evaluate non-probabilistic reliability of the example structure, finally non-probabilistic reliability index based on two reliability models is analyzed and compared, which reveal their difference and interrelation.Secondly, the non-probability reliability optimization is a nested optimization problem, outer loop is to optimize the target function of structural performance,which satisfy the non-probability reliability index; inner loop is a min-max problem associated with evaluation of the reliability index. To solve this problem, the outer loop adopting genetic algorithm to calculate, the inner loop using MATLAB optimization toolbox to solve, the numerical example shows that the method presented here is correct and feasible.Finally, non-probabilistic reliability optimal design of the main girder of the bridge crane metal structure is performed. The parametric finite model is established based on ANSYS-APDL language, checking the finite analysis results that stress and deflection; Using isight software to build response surface model of the crane girder structural strength and stiffness with the uncertain parameters and design variables, based on interval model to calculate the non-probability reliability index; With the volume of the main girder structure as the objective function, carry on the traditional structure optimization and the non-probabilistic reliability optimization, The results show that the results of traditional optimization did not meet the requirements of reliability, while the non-probabilistic reliability optimization results meet the design requirements and effectively reduce the volume of main girder structure.In this paper, considering the uncertainty of structural parameters, the crane girder structure is optimized based on response surface model and non-probabilistic reliability, the presented method for complex structure of the uncertain parameters optimization design has a certain reference value and significance.