薄板试样定向凝固枝晶竞争生长

It is of great scientific and engineering significance to understand the misoriented grain evolution in the directional solidification of superalloys, where the competitive growth of dendrite arrays with different orientation is the key phenomenon to understand the grain boundary evolution and elimination of dendrite. The experimental bicrystal growth in directional solidification identified the drawback of the traditional tip undercooling criterion. Moreover, quantitative phase field simulations showed more details of the bicrystal competitive growth and enhanced our understanding. In this research, we proposed a sheet sample with multiple bicrystals configurations. The microstructures were observed in a stitching picture. The results indicate that the competitive growth of two dendritic arrays is directly related to the interaction of primary arms and secondary arms in the grain boundary region. In the converging growth, the primary arm of the favorably oriented dendrite block the primary arm growth of unfavorably oriented dendrite and the new primary arm of the favorably oriented dendrite can not overgrowth in the grain boundary region. Then the grain boundary forms mainly along the growth direction of favorably oriented dendrite. In divergent growth, when the two dendritic arrays are on the same side of thermal gradient direction, the grain boundary is still mainly along the growth direction of favorably oriented dendrite where the new primary arm of unfavorably oriented dendrite is more likely to form than that of favorably oriented dendrite. When the two dendritic arrays are on the different sides of thermal gradient direction, the new primary arms of the both dendritic arrays can generate and the grain boundary is between the growth direction of the two arrays. The competitive growth and the grain boundary selection of dozens of biocrystals have been summarized within several samples. The statistical results agree with the phase field simulation. However, the CAFE simulation results from the ProCAST can not predict the grain boundary evolution due to there is only a profile of grains related to tip undercooling without solute diffusion and competition of side branches in CAFE. The method proposed here can be used to obtain the statistic information of the competitive growth with a high throughput way, and the results are helpful to understand the misoriented grains in fabrication of single crystal blade.