Whether at a blacksmith’s forge or in a modern steelworks, precise heat control is essential when shaping metals. Today, additive manufacturing (AM) processes such as laser powder bed fusion (L-PBF) use finely tuned lasers to create similarly controlled, localised thermal histories at the microscale, enabling the 3D-printing of complex shapes layer by layer from metal powders. This makes L-PBF a potentially time- and cost-saving method for producing critical components for aircraft, marine vessels and other heavy industries.
However, high-strength steels can be tough to print. “During L-PBF, the material goes through rapid cycles of melting, solidification and cooling, which make its final crystalline microstructure highly sensitive to printing conditions,” said Pan Wang, a Senior Scientist at the A*STAR Singapore Institute of Manufacturing Technology (A*STAR SIMTech). “In addition, typical strengthening approaches, such as cold working or thermomechanical processing, are difficult to apply to near-net-shape AM parts.”
But what if these repeated heating and cooling cycles could be turned into an in-built heat treatment? To forge a sharp yet impact-resistant knife, a skilled blacksmith applies different heating, quenching and tempering steps across the metal, hardening the edge while preserving toughness elsewhere. In a similar way, L-PBF could use laser-controlled thermal histories to tune the mechanical properties of different regions within the same component.
“With L-PBF, every small volume of material—only about as thick as two human hairs in cross-section—is heated repeatedly by neighbouring laser passes and successive layers of deposited material,” Wang explained. “This resembles an extremely fast and highly localised version of conventional quenching and tempering.”
Wang and A*STAR SIMTech colleagues worked with Jun Ding and colleagues at the National University of Singapore, as well as collaborators from Nanyang Technological University; Newcastle University, UK; and Xi’an University of Technology, China, to explore this possibility using AISI 4340, a high-strength steel commonly used in demanding load-bearing parts such as aircraft landing gear, heavy-duty gears and drive shafts.
Behind AISI 4340’s mechanical strength are two key crystalline phases: austenite and martensite. Austenite forms at high temperatures, while rapid cooling transforms it into harder and stronger martensite. During L-PBF, the first pass of a laser over a small region drives rapid phase transformation, converting much of the high-temperature austenite into martensite. Successive reheating by neighbouring tracks and later layers can then interrupt or reverse parts of this transformation pathway, leaving a small amount of retained austenite in the printed steel.
By combining experiments with thermal simulations, the researchers found that laser power and scanning speed determined how much retained austenite survived when printing AISI4340 with L-PBF. Simulations showed that lower-power and lower-scanning speed settings created a thermal history that preserved substantially more retained austenite, giving the printed steel a useful reserve of strengthening capacity.
“We found that when the steel was stretched, its retained austenite progressively transformed into fresh martensite,” said Wang. “As a result, the material does not simply start strong; it continues to strengthen while being deformed. This delays localised deformations that would otherwise lead to necking and fracture.”
By adjusting laser parameters, the team produced AISI 4340 parts with an ultimate tensile strength of 1,747 MPa with elongation over 10 percent without additional ageing or heat treatments, placing their outputs at the upper end of as-printed L-PBF high-strength steels.
“Our results show that thermal history controls retained austenite content, and therefore affects strain hardening and mechanical behaviour,” said Wang. “The next step is to move from simply understanding this relationship to deliberately programming it.”
The A*STAR-affiliated researchers contributing to this research are from the A*STAR Singapore Institute of Manufacturing Technology (A*STAR SIMTech).
