Since　the　surface　quality　of　the　NiTi　parts　prepared　by　selected　laser　melting　(SLM)　technology　is　generally　difficult　to　meet　the　requirements　of　the　application　due　to　the　high　surface　roughness,　the　work　aims　to　investigate　the　effects　of　laser　power,　scanning　speed　and　hatch　space　on　the　surface　roughness　of　NiTi　alloy　samples　during　the　process　of　SLM　NiTi　parts.　The　pre-alloy　NiTi　powders　were　atomized　by　the　electrode　induction　melting　gas　atomization　technique　(ALD,　Germany)　under　the　protection　by　argon　gas.　NiTi　powders　were　observed　by　scanning　electron　microscopy　(SEM).　The　particle　size　distribution　(PSD)　was　measured　by　laser　scattering　particle　size　analyzer　(HORIBA　LA-960S,　Japan).　For　experimental　design,　the　model　was　created　based　on　the　Taguchi　design　model　of　the　L16　orthogonal　array.　The　range　of　laser　power,　scanning　speed　and　hatch　spacing　was　20-50　W,　200-500　mm/s　and　0.05-0.08　μm.　The　scanning　strategy　of　laser　rotation　67°　between　two　consecutive　layers　was　applied　to　produce　NiTi　parts,　with　a　fixed　layer　thickness　of　30　μm.　NiTi　samples　of　6　mm×　6　mm×　6　mm　were　produced　by　SLM　technology　(EOS　M100,　Germany).　The　surface　roughness　value　and　surface　morphology　were　measured　by　laser　confocal　microscope　(Olympus　LEXT　OSLS4100,　Japan).　The　surface　roughness　signal-to-noise　(S/N)　ratio　was　calculated　by　the　equation.　The　S/N　ratio　was　a　logarithmic　function　used　as　an　objective　function　for　optimization,　which　was　conductive　to　data　analysis　and　prediction　of　optimal　results.　Samples　are　successfully　prepared　by　SLM　technology.　The　results　showed　that,　at　the　laser　power　of　20　W　and　30　W,　the　sample　surface　had　high　fluctuation　due　to　the　powder,　which　could　not　melt　sufficiently　to　the　unstable　melt　track　during　the　process　of　SLM　NiTi　parts,　with　a　maximum　surface　roughness　value　of　7.8　μm.　When　the　value　of　laser　power　reached　50　W,　the　sample　with　low　surface　roughness　value　was　obtained　which　was　attributed　to　the　stable　melt　track,　with　a　minimum　surface　roughness　value　of　1.3　μm.　The　sample　surface　roughness　value　increased　with　the　increase　of　scanning　speed,　at　the　same　laser　power,　which　was　attributed　to　the　time　of　NiTi　powder　melting　increasing　at　low　scanning　speed.　The　decreasing　of　hatch　spacing　could　remelt　the　adjacent　laser　track　to　improve　the　surface　morphology　of　NiTi　parts.　However,　the　surface　roughness　affected　was　not　obvious　when　the　laser　track　was　not　unstable.　The　rank　order　of　the　process　parameters　on　the　surface　roughness　is　laser　power,　scanning　speed,　hatching　spacing　after　statistical　methods　are　used　to　analyze　the　surface　roughness　signal-to-noise　ratio.　According　to　the　mathematical　model,　the　optimal　combination　of　process　parameters　is　laser　power　of　50　W,　scanning　speed　of　200　mm/s　and　hatch　space　of　0.07　mm.　The　surface　roughness　value　(1.38　μm)　of　the　sample　prepared　by　the　final　optimized　process　parameters　combination　is　close　to　the　predicted　value　(1.43　μm)　by　fitting　equation.　The　difference　of　the　surface　roughness　value　is　only　9.97%.　The　model　provides　an　accurate　guide　for　experiments　for　the　study　of　SLM　NiTi　parts.　©　2024　Chongqing　Wujiu　Periodicals　Press.　All　rights　reserved.