Wind Resistant Optimization Of Double-layer Cylindrical Reticulated Shell Based On The Optimal Criterion

With the development of society, large-span roof structures have been applied widely in public facilities. This kind of structure is very sensitive to wind load and the traditional design for them always result in the waste of materials due to conservative mind so that we need to study the wind resistant optimization design on them. However, currently the researches on wind resistant optimization design mainly concentrate on the high-rise buildings. Therefore, it is necessary to carry out the study on the wind resistance optimization of large-span roof structures. At the same time, the Optimal Criterion method, referred to as the OC method, has been highly used in the optimization design of high-rise structures; nevertheless, its application in large-span roof structures is relatively few. On the other hand, the previous studies of wind-induced optimization with OC method on large-span roof structures have considered a small amount of constraints. This work has considered four types of constraints(namely, tension constraint, slenderness ratio constraint, vertical displacement constraint and stability constraint) and included 247 numbers of constraints totally. Programming by Matlab with the Harmonic Excitation Method referred to as HEM, the Load Response Correlation method referred to as LRC and the OC method, this study has conducted wind resistant optimization design to a double-layer cylindrical reticulated shell structure. The following conclusions are obtained:(1)The number of iterations required for each design in this dissertation is less than 10,and all of the designs have achieved the economical and safe structure. Hence,these design results have proved that the OC method is applicable to large span roof structures as well,having fast convergence and less impacted by the design variable numbers and structural scale.(2)In order to releasing and eliminating the fake convergences happenning in the design process, it is benefit for us to timely increase the design variable updating step, narrow convergence control value, increase the number of constraints, and use other ways in place of the direct derivation method to calculate sensitivities.(3) Sensitivity analysis clearly indicates that the objective function and constraint functions are influenced differently by each design variable and these sensitivities may change in design process. Combining these features with the information of Lagrange multipliers, we can predict the tendency of optimization and thereby understand the dominating design variables or structural performances, which could help us improve the design efficiency.(4)According to that the eigenvalues of Hessian matrices of objective function and constraint functions are larger than or equat to zero, it is proved that they are convex functions and the optimization problems are convex programming problems. So, the results acquired in this work are the optimal solutions.(5)Even if only the slenderness ratio and stability constraints are considered, the expected target can be achieved, and the optimized structure also satifies the demands related to specifications and constraints on stress and displacement.(6)The natural frequencies of structures are one of the most important parameters in their wind resistant optimization design, so the uncertainty of structural parameters in real engineering is tentatively explored by using Interval Factor Method referred to as IFM. The dynamic characteristics of a double layer cylindrical reticulated shell structure under certainty and uncertainty situations are analyzed. The relationship between the interval dynamic characteristics and the uncertain structural parameters are understood. Moreover,according to that the interval factor has the identical changing rate among different orders of frequencies or modes, any order of interval natural frequencies or modes can be calculated.