Resarch on Microstructure and Property of Zirconium alloy in the Rotary forging process by GNN-FEM

Abstract

In this paper, the effect of resarch on rotary forging process of Zirconium alloy bar was studied by combining Metallographic observation, electron microscope observation, Convolutional neural networks (CNN) and finite element (FEM). The results show that Point pits and scratches often occur in rotary forging. Modifying die size and increasing entrance Angle can effectively improve such defects as point pits and scratches. The external structure of Zirconium alloy bar is relatively dense, and the internal structure is relatively loose. The number and size of micro-holes after acid erosion are small. The number and size of the tiny holes after internal acid erosion are large. The (0002) planes of the inner and outer grains of the Zirconium alloy bar are basically parallel to the rolling direction, but the outer (0002) is concentrated on the pole, and the inner (0002) pole is approximately distributed in a ring around the longitudinal axis. As a result, there are differences in the small holes inside and outside the bar after corrosion The effect of resarch on the classification Method of RF in by GNN, which the Accuracy-Avg is 96%. At the beginning of RF, the rolled pass exerts compressive stress on the outside of the Zirconium alloy bar, and the outside of the bar is the first to undergo plastic deformation. With the increase of RE deformation, the bar strain gradually increases and is gradually transferred from the outside to the inside. The effect of resarch on the classification Method of by GNN-FEM is an efficient and convenient method to simulate the microstructure and stress distribution of zirconium alloy in the process of RF.