Study on milling surface integrity and tool wear of Polysynthetic twinned TiAl single-crystal alloy based on acoustic emission signal

To address difficult machining and unclear cutting mechanism of Polysynthetic twinned (PST) TiAl single-crystal alloys, this study carries out milling tests to investigate surface integrity and tool wear. The relationship between acoustic emission (AE) signals, surface integrity and tool wear is analyzed. Based on this, tool wear mechanism and machined surface generation mechanism are clarified. Combined with tool wear state, the variation mechanism of AE signal characteristic values is studied, and a prediction model of AE signal eigenvalues with surface integrity and tool wear state is established. Aiming at the robustness of the experiment, multiple experiments are carried out and the wear state of all cutting edges is measured. The results show that the initial wear stage limit of three kinds of cutting tools is approximately 130 mm³, and the corresponding wear widths are 20.1 μm, 11.2 μm and 8.4 μm respectively. When the material removal volume exceeds 130 mm³, the multi-layer coated tool exhibited delamination and peeling, and bending degree of the lamellar structure is deepened. Meanwhile, the frequency of the center of gravity of the AE signal, the range and average value of the frequency variance decrease. Then, milling force, plastic deformation layer depth and machined surface roughness increase by an average of 30%, 65% and 30% respectively. Meanwhile, deformation angle of subsurface lamellar structure increases. Then, obvious milling marks, involving wave crests, and pits appear on workpiece surface. PST TiAl single-crystal alloy workpiece microhardness increases by 6.3%, and machining softening occurs in the depth range of 10-40 μm. Mathematical models of AE signals with tool wear, milling force, plastic deformation layer depth and surface roughness are established, and the sensitivity coefficients of different AE signal eigenvalues are calculated. Then, sensitivity coefficient of average root-mean-square frequency is the highest.