Printing-Parameter Effects in SLM 316 L Stainless Steel

Metal additive manufacturing is increasingly being examined for structural engineering use, where mechanical reliability depends on strength as well as repeatability, ductility, anisotropy, and compatibility with established constitutive descriptions. Selective laser melting is especially relevant for stainless steel components because it can produce geometrically complex parts through layer-by-layer melting of metallic powder. That manufacturing route, however, gives the material a process history very different from conventional forming. Each layer experiences rapid heating, melting, cooling, and solidification, and the final material response is shaped by the interaction of melt pool geometry, interlayer bonding, scan path, build orientation, and local defects. For constructional steel applications, these details cannot be treated as secondary manufacturing features; they become part of the material definition itself.

The main challenge is that SLM 316 L stainless steel does not have a single mechanical identity independent of how it is printed. Earlier studies had examined individual process variables such as laser power, scanning speed, scanning spacing, layer thickness, scanning strategy, specimen thickness, forming method, and printing angle. However, the available experimental basis remained limited when several practical parameters were considered together. This matters because structural design requires material models and design parameters that can represent the actual stress-strain response, not only isolated strength values obtained under one manufacturing condition. A stainless steel part printed at one angle or with one scanning strategy may satisfy strength expectations while showing a different fracture strain, strain-hardening behavior, or directional response from a part produced under another condition.

In a recent research paper published in Journal of Constructional Steel Research, Professor Zhi-Wei Shan from Southeast University working together with Professor Bing-Bing San, Dr. Fang Xu, and Dr. Wen-Hui Zhao from Hohai University developed an experimental dataset for SLM 316 L stainless steel tensile behavior across combined scanning strategies, specimen thicknesses, forming methods, and printing angles. They also evaluated modified two-stage Ramberg-Osgood constitutive descriptions against the measured stress-strain curves and identified the Gardner and Ashraf formulation as the preferred model for this material. They also identified suitable formulas for predicting strain-hardening exponents, ultimate stress, and ultimate strain in the tested material.

The researchers prepared 42 tensile specimens from 316 L stainless steel powder by selective laser melting, arranged across 21 groups with duplicate specimens to improve data reliability. Their experimental design included three scanning strategies, four specimen thicknesses, two forming methods, and three printing angles. The scanning strategies were bidirectional scanning, checkerboard scanning, and bidirectional scanning with 67° interlayer rotation. The forming routes distinguished directly printed specimens from specimens cut from rectangular plates, while the printing angle described the angle between the tensile loading direction and the laser beam reciprocation direction. This design choice mattered scientifically because it allowed the tensile response to be separated into manufacturing-related contributions rather than reduced to a single average material curve. The stress-strain behavior resembled that of conventional stainless steel in one important respect: the curves did not show a distinct yield plateau. However, the mechanical response was strongly shaped by printing direction and processing route. Across the tested specimens, elastic modulus, yield stress, ultimate stress, and fracture strain covered broad ranges, from 164 to 206.9 GPa, 370 to 532 MPa, 539 to 685 MPa, and 16.8% to 68.2%, respectively. These ranges make clear that SLM 316 L stainless steel cannot be described adequately by one nominal tensile value when the printing parameters vary.

The team scanning strategy had a marked effect on strength but a weaker effect on elastic modulus. They found that specimens printed with bidirectional scanning and 67° interlayer rotation showed higher yield and ultimate stresses than those printed with the other two scanning strategies, while their fracture strain was lower. The elastic modulus remained nearly insensitive to scanning strategy. Thickness produced a different pattern: yield stress and ultimate stress increased as specimen thickness increased, while elastic modulus and fracture strain for 4 mm and 5 mm specimens were broadly similar to those for 2 mm and 3 mm specimens. Forming method had little influence on elastic modulus, yield stress, and ultimate stress, but it affected fracture strain, with cut specimens being more susceptible to premature fracture because of internal and surface defects introduced or exposed by the cutting route.

The authors performed printing angle studies and showed the 45° specimens generally showed higher elastic modulus, yield stress, and ultimate stress than the 0° and 90° specimens, while fracture strain increased as the printing angle increased. The fracture observations were consistent with this mechanical trend. Specimens printed at 0° showed sudden fracture without obvious necking, whereas 90° specimens displayed better ductility and a clear necking phenomenon. The link between printing angle and ductility was interpreted through the relation between loading direction and interlayer shear direction: when the interlayer shear direction was parallel to the tensile loading direction, ductility was reduced; when it was perpendicular, ductility improved.

Afterwards, the researchers examined modified two-stage Ramberg-Osgood descriptions for the stress-strain response of SLM 316 L stainless steel. They observed the Gardner and Ashraf model matched the tested curves more across the full strain range than the Rasmussen formulation, particularly for 0° and 45° specimens, where the Rasmussen model tended to overpredict the strain-hardening stage. For material parameter prediction, the two-point formula based on σ0.05 and σ0.2 gave better accuracy for the first strain-hardening exponent, while the EN 1993-1-4 expression was more suitable for the second strain-hardening exponent. Existing formulas for ultimate stress also performed reasonably for SLM 316 L stainless steel, and the ultimate strain prediction remained conservative, partly because the material combined a high yield ratio with a strong printing-angle effect.

The findings of the study are directly relevant to civil and constructional steel applications where selective laser melting can produce complex stainless steel components, customized joints, and geometrically efficient elements whose performance depends on geometry, load direction, and fabrication route.   One important application is in the design of printed stainless steel connectors, nodes, brackets, and transition pieces for truss or cable-supported structures. Specimens printed at 45° achieved higher elastic modulus, yield stress, and ultimate stress than 0° and 90° specimens, while fracture strain increased with printing angle. This means that engineers can use build orientation as a design variable, not just a manufacturing convenience. When strength is the primary concern, a 45° orientation may be advantageous; when ductility and deformation capacity are more critical, the improved fracture strain at higher printing angles becomes important. The results also support process selection for load-bearing SLM stainless steel parts. The bidirectional scanning strategy with 67° interlayer rotation produced higher yield and ultimate stresses than the other scanning strategies, although with lower fracture strain. This has practical value for components where load-bearing resistance is prioritized while also clarifying the associated strength–ductility balance. Thickness effects are also useful for structural detailing, since yield and ultimate stresses increased as specimen thickness increased, whereas elastic modulus and fracture strain were less strongly altered across the tested range. The new work evaluates modified two-stage Ramberg-Osgood models and recommends the Gardner and Ashraf formulation for representing the stress-strain behavior of SLM 316 L stainless steel. This gives engineers a more reliable constitutive description for finite element analysis, nonlinear structural assessment, and performance-based design of printed stainless-steel elements.