The effect of process parameters and scanning strategies on surface roughness of stainless steel 316L SLM parts

Abstract

Selective Laser Melting (SLM) process is a powder-based metal Additive Manufacturing (AM) technique that can generate fully functional and geometrically complex shapes which can be used in actual load-bearing situations. The mechanical properties of SLM parts are highly dependent on process parameters. Generally, the as-built surface quality of SLM parts is not good enough, and finding the optimum values of the parameters to improve surface quality is an ongoing challenge. In this study, an experimental approach was taken to investigate the effect of process parameters and scanning strategy on the surface roughness of stainless steel 316L parts as well as determine the optimal values of parameters and optimum scanning strategy. Laser profilometry was applied to measure surface roughness. Results showed that the scanning speed was the most effective parameter; increasing it from 900 mm/s to 1200 mm/s caused 62.5% rougher surfaces. Also, increasing laser power from 255 W to 300 W resulted in 41.8% rougher surfaces. Smaller hatch spaces caused better surface quality; decreasing it from 0.1 mm to 0.04 mm improved surface quality by 16%. And lower scan pattern angles resulted in slightly rougher surfaces (by 2%). Changing the settings from minimum overlap to maximum overlap resulted in 73.6% better surface quality. It was found that insufficient melting is the main reason for poor surface quality. Surface quality was improved whenever the resultant effect of parameters caused sufficient melting conditions. The optimum values were found as 270 W laser power, 900 mm/s scanning speed, 0.06 mm hatch spacing and 75° scan pattern angle. The best surface quality was obtained for the chessboard scanning strategy, with scan vectors being parallel to each other in adjacent blocks and rotating by 90° in successive layers. The optimum strategy caused 2.5 times better surface quality than the strategy of the roughest sample.