Thermal Analysis Methods For Ceramic Matrix Composite Turbine Vanes

Continuous fiber reinforced ceramic matrix composite(CMC) has become the most competitive alternative material for the future advanced aero-engine due to its small density and good high-temperature resistance. This thesis aims to investigate the thermal analysis method for CMC turbine vanes. Innovative method of predicting equivalent thermal conductivity(ETC) based on microstructure image recognition has been established, then the anisotropic thermal conductivity and its disparity were taken into account in the numerical simulation of temperature, finally a thermal analysis method for CMC turbine vanes was completed together with the custom program.In this thesis, typical continuous unidirectional fiber reinforced composite was firstly applied as the object of study. Scanning Electron Microscope(SEM) was used to get the microstructure and the ETC prediction method based on microstructure image recognition was investigated. Then the ETC of T300 carbon fiber/epoxy resin unidirectional composites was calculated by using this new method. The relative error of numerical prediction against experimental data is 10.1% and the absolute error is only 0.035 W/(m?K). With the evaluation of this method, further studies about the influence of gaps among fibers and matrix on ETC were carried out by it. The results show that the ETC of unidirectional composites reduced gradually with the increasing of thickness and proportion of the gaps. The randomness of gap position affects ETC weakly when the proportion of gap remains constant.Then three-dimensional five-directional braided composite was applied as the object of study. Representative volume element(RVE) was established also by the method based on microstructure image recognition. The ETC of a three-dimensional five-directional braided SiCf/SiC composite sample was calculated, considering the fiber yarns' arrangement patterns and spatial extend direction. By comparing the prediction results with experiment data, the relative error is only 5.49%. Then the effects of structure parameters, thermal properties of components and structure parameters' disparities on the ETC were studied in detail. It was found that both the vertical and horizontal ETC were enhanced with the increasing of the fiber fill factor. The results also indicate that the dispersion of fiber fill factor affects ETC much more than the internal braiding angle's dispersion.Based on these studies on the anisotropic and dispersive characteristics of ETC, a thermal analysis software for air cooled CMC turbine vane was developed by programming with Fortran and Matlab. Many useful functions are integrated in this platform, including ETC prediction for composite, external and internal heat transfer boundary setting, temperature field simulation and parametric modeling for cooling configuration.Finally, three-dimensional five-directional braided SiCf/SiC composite turbine vane model was studied with this software platform. The distribution of temperature was calculated and it was found that the potential high temperature area extended into solid region much more if the internal braiding angle's dispersion was considered. The total area of potential high temperature region increased by 13.6%. The numerical results show that the distribution of high temperature area is mainly determined by the spatial characteristics of ETC, while the temperature values of those regions are closely related to the degree of ETC's anisotropy.