| The natural draught cooling tower is widely used in thermal power stations and other industrial fields. The wind load is the dominant load of cooling tower due to its large size, complex geometry and thin-wall structure. Currently, the wind pressure distribution is generally employed in the wind load effect assessment for the large-scale cooling tower, which is based on code regulations, wind tunnel test or computational fluid dynamic (CFD) analysis. In practice, these methods neglect the structural deformation effect on the wind load, which can cause irrational results.Firstly, the multiphysics analysis method based on the iterative pressure has been used to study the response of the large hyperbolic cooling tower under steady wind load, and the interaction between cooling towers subjected to wind load has been studied. It is found that the code-based method is conservative, and the wind load on the outer surface of the tower is dominated, and the front tower has much more influence on the behind tower. Secondly, the method of Fluid-Solid Interaction (FSI) has been used for structure analysis under the unsteady wind load. For the cooling tower under the unsteady wind load, the transient responses in the initial several seconds are relatively stronger.On the premise of ensuring the accuracy, how to improve the efficiency of the analysis is the long-term challenges for the complex structures. The structural reduction method is always one of the important means in improving the computational efficiency. The applicability of the structural approximate model has been performed in this paper. Two weak FSI models, based on the structural approximate models by the rigid-beam element and the flexibility modification, have been developed to improve the efficiency of structure dynamic analysis. The results of3D cylindrical structure under fluid load show that results of the weak FSI are comparable for the two structural approximate models. The difference will increase regardless of the interaction between fluid and structure. In general, the results of approximate models proposed in this paper are comparable with the resulits of ANSYS FSI elastic model. Moreover, differences by the fluid models of different fluid solvers require further study. For the structural responses, the top maximum across-wind displacement will increase nonlinearly with the increase of wind speed.The optimization is an iterative process, so the computation efficiency is key to the large scale and complex structure. For the large hyperbolic cooling tower under steady wind load, a 2-level optimization strategy has been developed in this paper, which employs the design of experiments (DOE) to build up a numerical approximation of complex structure. Firstly, Multi-island genetic algorithm is used for global optimization and then the sequential quadratic programming-NLPQL technique is used to obtain more satisfying optimal results. During the optimization, the preliminary results can be obtained quickly by the effective code-based method in the1st level optimization, based on which, more satisfying optimal results are assessed by the iterative method in the2nd level optimization. At the same time, the design space can be reduced properly and rebuilt near the1st level optimal results.Finally, based on the proposed approximate model, the local nonlinear problem has been investigated in this paper. It is generally known that the approximate model can improve the structural computational efficiency, but the traditional approximate model can’t handle the local nonlinear problem directly. In this paper, the approximate model based on flexibility modification is used as a fast solution method in view of local nonlinear dampers. The analysis results demonstrate that the proposed method has high accuracy and efficiency, which can consider the effects of damper failure on the structural responses. It is found that the proper control strategy for damper designs can protect the primary components of structures. |
PDF Full Text Request