Research On Parametric Block-Decomposition Methods Of Turbine Cooling Blades’ Mesh-Generation

It is important for turbine blades in aircraft engine to transform heat energy into mechanical energy. The blades work in the condition of high temperature and high pressure, which will lead to the complex boundary conditions, irregular blade-shape and plentiful cooling structures. All these will make a heavy work when it comes to modeling, block-decomposition and mesh generation.This paper, was written to analyze the characteristics of blades’ structure, to define model-expression with basic geometric elements by Boolean methods, and to correspond geometry with topology. According to the geometrical characteristics, a universal block-decomposition method is applied concerning physical properties and boundary conditions. This method makes the computational domain-decomposition strategy into template integrating in software. After the decomposition, structured meshes should be generated with the step-size by TFI. In order to solve the possible problem that grid nodes are not corresponding between blocks, the adjacent surface should be treated as an interface. All the above, will make it possible to realize blades’ parametric modeling, automatic block-decomposition and structured mesh-generation through software program.Three aspects are involved in automatic grid-generation of turbine cooling blades. Parametric modeling. By analyzing the xml-file of the designed blade, the complex three-dimensional shape is ultimately be accurately described. And users can control the physical geometry structure adapting to the characters’ parameters. Block decomposition. Base on the shape of blades, the whole domain is parted into three pieces- the external flow of hot-gas, the internal flow of cooling-air, and the solid area between hot-gas and cooling-air. Each smaller piece of them should meet the conditions of structured- mesh generation. TFI algorithm for structured grid generation is used to achieve the structural meshing of the computational domain. The grid parameters are under control, and the mesh data will be saved after generation.