Fig. 1. (a) ECCI results of the oxide inclusion distribution (a typical region with a big inclusion), (b) the inclusion size distribution and the cumulative percentage of the inclusion size in bulk SLMed 316 L SS, (c) (d) HRTEM micrograph showed the interface between the inclusion and surrounding matrix, (e) Scheil model simulation: temperature as a function of the fraction of solid for 316 L SS composition with 0.03 wt.% oxygen content.

Fig. 2. (a) IPF + IQ maps and (b) grain size distribution of SLMed 316 L SS on the plane vertical to the building direction, (c) non-uniform cellular structure distributes in the etched SLMed 316 L SS surface (the plane is vertical to building direction) and the white arrows indicate the big cellular structure.

Fig. 3. (a) Engineering strain-stress curve and (b) true strain-stress curves and strain hardening rate curves for three SLMed 316 L SSs.

Fig. 4. In-situ ECCI results focusing on a typical region that contains a big inclusion: (a) no deformation, (b1) (b2) (b3) the cell boundary distance along the slip direction at different regions marked in (a), (c) 2% deformation. An enlarged area for ECCI observation (d) before and (e) after 2% deformation. The white arrows indicate the dislocations, the green arrows show the slip bands, and the dashed red arrows show the dislocation moving directions, respectively.

Fig. 5. In-situ tensile test at the big inclusion region: (a) engineering strain and stress curve during the in-situ tensile test at different strain levels, (b1-b7) a series of secondary electronic images of the big inclusion evolution under different deformation levels until fracture.

Fig. 6. (a1-a7) Enlarged secondary electronic images of the big and small inclusions under different deformation levels until failure. The left and right crack length near the inclusions with the global engineering strain: (b) big inclusion (～500 nm) and (c) small inclusion (～80 nm), respectively.

Fig. 7. EBSD results for the in-situ tensile test under different deformation levels. For 15 % deformation level: (a1) IPF + IQ maps, (a2) KAM map, (a3) misorientation along the red line. EBSD result of the polished surface after the failure for the SLMed 316 L SS: (b1) IPF map, (b2) KAM map and (b3) ECCI results of the same area, showing that the deformed nano-twins can grow to cross the small inclusions. For 25 % deformation level: (c1) IPF + IQ maps, (c2) KAM map, (c3) ECCI results of the deformed nano twin cluster, white arrows indicated the nano-twin sites. The sketch diagram for the dislocation, slip, crack and nano twin evolution with the increasing strain levels: (d1) initial stage, (d2) slip band initiation and crack occurred near the inclusion, (d3) nano twin initiation (d4) nano twin growth and crack propagated, respectively. The big inclusion sites are marked using big red arrows in (a1, a2, c1, c2).

Fig. 8. (a) The secondary electronic image of the tensile sample after failure with the equal cracking length for the left and right part of one big inclusion, (b) Statistical percentage results of the symmetric and asymmetric cracking on the sample surface after failure, more than 20 cracks were taken into accountMeanwhile, even cracks are first formed around the big inclusion; the earlier formed deformation-induced nano-twins around the inclusion can resist the crack propagation, which results in a very high localized plasticity around the inclusion and maintains an overall high post-necking elongation for the bulk materials. In the future, we could design the cellular structure configuration around the inclusions by changing the printing parameters, to control the cracking path. If a tortuous cracking path can be achieved, then a tough bulk AM material can be expected. However, the crack damage caused by big inclusion is much larger than the small inclusion, indicating the modest detrimental effect on the global toughness for small inclusions. Further reducing the size and density of the oxide inclusions in additively manufactured components is needed, methods, such as reducing the oxygen content both for the powder and printing atmosphere or adding some high oxygen consumption alloying elements, can be conducted to obtain better mechanical properties.