Thermal behavior and control during multi-track laser powder bed fusion of 316 L stainless steel

Understanding and controlling the thermal behavior during laser powder bed fusion (LPBF) additive manufacturing (AM) is crucial to improve the quality of the printed layer and the performance of the final part. In this paper, the computational fluid dynamics (CFD) method was used to numerically reproduce the multi-track LPBF process of 316 L stainless steel. The influences of key operating parameters including laser power P, scanning velocity V, and hatch spacing H on the heat affected angle, depth and related defects of multi-track molten layers were systematically analyzed, based on which a thermal control map for practical printing and manufacturing was constructed. In addition, the back propagation neural network (BPNN) model was utilized to predict the performance of molten layers under any combination of conditions. Results show that in the multi-track printing, the previous molten layer served as the preheated ‘substrate’ and powder layer, resulting in the increase of temperature and size of the next molten layer. Increasing laser power and decreasing scanning velocity/hatch spacing would induce the increase of heat affected angle and molten pool depth due to the large energy input and strong thermal interaction. It is found that when P = 240 W, V= 0.4 m/s, H= 60 µm, the maximum heat affected angle and molten pool depth are 17.3° and 130.93 µm, respectively. The thermal interaction between molten layers could be reduced by lowering the laser power or increasing the scanning velocity and the hatch spacing. However, it should be noted that near the critical point (P = 160 W, V=1.2 m/s, H=140 µm), the defects such as insufficient or non-melting and pores were prone to form in the molten layer. The multi-track molten layer with low thermal effect and high molten pool depth could be easily obtained by the thermal control map. The BPNN model could well predict the values of heat affected angle and molten pool depth, providing a convenient way to obtain the more precise thermal control map. The novel findings from this research will provide valuable references for obtaining qualified molten layers as well as high performance parts with desired precision.