Selective laser melting builds parts layer by layer, tracing each slice of a digital model with a laser that melts fine metal powder in its path. The result is a material with an internal memory of its own construction. Melt pools solidify into elongated grains, layer boundaries trace horizontal planes, and melt channels record the laser’s scan paths. In 316L stainless steel, an austenitic alloy prized for its corrosion resistance and weldability, these features give rise to pronounced anisotropy: mechanical properties measured along the build direction differ measurably from those measured perpendicular to it. Engineers have long accounted for this in tensile and fatigue design. What remained poorly understood, the authors note, is whether stress corrosion cracking, the marriage of sustained tensile stress and a corrosive environment that can fell structures without warning, follows the same directional rules, and whether those rules hold as exposure stretches from seconds to weeks.

To answer that question, the team designed a three-pronged experimental campaign spanning three distinct timescales. Samples were printed in two orientations: one loaded along the build direction, designated the BD sample, and one loaded vertically to the build direction, the VBD sample. The shortest test was an ultra-short-term electrochemical corrosion test, which probes the earliest moments of interaction between the metal surface and a chloride-bearing electrolyte. The intermediate test was a 48-hour slow strain rate stress corrosion test, in which specimens are pulled apart slowly while immersed, allowing corrosion and deformation to interact. The longest was a 720-hour constant load stress corrosion test, a full month of sustained stress in a corrosive medium that mimics the slow, patient degradation of real service. This triad of methods, rarely combined in a single study, allowed the researchers to watch the same anisotropy express itself at every stage of damage evolution.

The electrochemical results set the tone. The BD sample exhibited a lower corrosion current density, meaning it corroded more slowly under applied potential, along with a larger pitting overpotential and a higher polarization resistance. In practical terms, the passive oxide film that protects stainless steel formed more stubbornly on the surface loaded along the build direction, and it took a greater electrochemical push to nucleate the first pits. Pit initiation, the critical first step on the road to stress corrosion cracking, was therefore delayed in the BD orientation. Because pitting corrosion is widely regarded as the gateway to cracking, since pits concentrate stress and create local chemistry that dissolves the passive film, this early advantage compounds over time. The finding aligns with a growing body of work showing that build orientation reshapes the stability of passive films on additively manufactured 316L.

The 48-hour slow strain rate tests translated that electrochemical head start into mechanical terms. The researchers quantified stress corrosion susceptibility using sensitivity indices computed from tensile strength, fracture elongation and fracture energy, comparing behavior in a corrosive environment against behavior in pure water. Across all three metrics, the BD sample proved less sensitive to the corrosive environment than the VBD sample. More revealing was where the damage began. In the BD orientation, corrosion pits and cracks initiated at printing layer boundaries, but their propagation was hindered, as if the layered microstructure acted as a series of speed bumps. In the VBD orientation, pits and cracks started at melt channel boundaries and then propagated linearly, carving relatively unobstructed paths through the material. The geometry of the microstructure, in other words, determines not just where cracks start but how easily they travel.

The month-long constant load experiments added the dimension of time-dependent deformation. Over 720 hours, the BD sample accumulated greater long-term deformation than the VBD sample, a seemingly counterintuitive result given its superior cracking resistance. But the damage mechanisms told a coherent story. On the surface of the BD sample, micro-cracks and corrosion promoted each other in a localized feedback loop, with each pit seeding a crack and each crack channeling corrosive species deeper. On the VBD sample, by contrast, corrosion partnered with large-scale plastic deformation, and their interaction produced voids that coalesced into more damaging failure. The BD material deformed more but failed more gracefully; the VBD material deformed less but failed more catastrophically. The overall verdict was unambiguous: the stress corrosion cracking resistance of the BD sample is superior to that of the VBD sample on all timescales tested.

To unify these observations, the team constructed anisotropic dynamic evolution diagrams, schematic maps that trace how damage accumulates differently in the two orientations as time unfolds. These diagrams capture a conceptual shift that is increasingly important in additive manufacturing research: anisotropy is not a static property that can be captured by a single tensile test, but a dynamic one that evolves as pits nucleate, cracks initiate, and microstructures either block or channel their growth. A part that looks adequate after a quick electrochemical screening may behave very differently after a month under load, and the study demonstrates that the ranking of orientations can only be trusted if it holds across the full spectrum of timescales. Here, fortunately for designers, the ranking did hold, giving engineers a rare degree of confidence in one orientation over the other.

The implications reach well beyond the laboratory. 316L stainless steel made by selective laser melting is already deployed in marine environments, chemical processing plants, biomedical implants and energy infrastructure, all settings where chloride ions, sustained stress and long service lives conspire to threaten stress corrosion cracking. The study suggests that print orientation should be treated as a first-order design variable for corrosion performance, not merely a geometric constraint. Components expected to endure corrosive service could be oriented on the build plate so that critical stresses align with the more resistant direction, and post-processing or scan strategy adjustments could target the melt channel boundaries that serve as crack highways in the weaker orientation. The work also provides theoretical support for qualification programs, which increasingly demand that additively manufactured parts demonstrate durability under realistic, long-duration exposure rather than accelerated proxies alone.

The study also slots into a broader scientific conversation about defects in printed metals. Prior research has tied pitting in SLM 316L to gas pores, residual stresses and surface finishes, and other teams have documented build-direction effects on passive film stability and cracking in high-temperature water. What distinguishes the new work is its systematic coupling of timescale with orientation, and its demonstration that the same microstructural features, layer boundaries and melt channel boundaries, play opposite roles depending on how the load is applied relative to them. As additive manufacturing matures from prototyping toward certified structural use, this kind of mechanistic clarity is exactly what standards bodies and safety engineers need. The message for the industry is quietly revolutionary: the fight against corrosion in 3D-printed steel may be won or lost before the part ever leaves the printer, in the invisible architecture of its layers.

Subject of Research: Anisotropic stress corrosion cracking behavior of selective laser melted 316L stainless steel across different timescales

Article Title: Anisotropy of stress corrosion cracking across different time scales for SLM 316L stainless steel

Article References: Hua, X., Lu, D., Miao, X., Wu, W., Dai, Q., & Peng, J. (2026). Anisotropy of stress corrosion cracking across different time scales for SLM 316L stainless steel. Journal of Materials Science. https://doi.org/10.1007/s10853-026-13683-y

Image Credits: AI Generated

DOI: 10.1007/s10853-026-13683-y

Keywords: 316L stainless steel, selective laser melting, stress corrosion cracking, anisotropy, additive manufacturing, pitting corrosion, build direction, electrochemical corrosion, slow strain rate testing, constant load testing, passive film, corrosion science

The research, led by Fatemeh Mehdipour of the Karlsruhe Institute of Technology together with colleagues including Thomas Koch and Roland Dittmeyer, set out to answer a deceptively simple question: can the design freedom of metal 3D printing produce a catalyst support that converts more pollution without choking the engine? Conventional ceramic monoliths are thermally robust but brittle, and the extrusion process that makes them restricts their interiors to simple square or hexagonal channels. That geometry produces largely laminar flow, modest heat and mass transfer, and susceptibility to maldistribution, all of which limit how efficiently harmful molecules reach the catalytic surface. Cold-start conditions, before the catalyst reaches its light-off temperature, compound the problem.

The team selected three TPMS architectures as the building blocks of their experimental carriers: the Schwarz-primitive, the gyroid, and the Schwarz-diamond surfaces. These are mathematical minimal surfaces that repeat periodically in three dimensions, producing smooth, continuous, fully interconnected channel networks with no sharp junctions. Such geometries can pack a large surface area into a small volume while promoting uniform flow distribution, offering a way to soften the classic trade-off between catalytic surface area and pressure drop. Each monolith was designed as a cylinder 50 millimetres tall and 23.88 millimetres across, sized to slot into a sample holder on an engine test bench, with unit cell size, orientation, and even axial size gradients varied systematically to isolate each design parameter’s influence.

Twenty-seven monoliths were fabricated from gas-atomised 316L stainless steel powder using laser powder bed fusion, with 0.2 millimetre walls guided by prior work showing that gas-tight thin walls in the 200 to 300 micrometre range are achievable with the process. Post-processing was deliberately gentle: ultrasonic cleaning flushed loose powder from the internal channels, and wire electrical discharge machining, rather than force-based cutting, separated the parts from the build plate to avoid distorting the thin-walled circular cross-sections. The finished samples measured within a few hundredths of a millimetre of their nominal dimensions, confirming that the intricate internal lattices could be manufactured reproducibly.

Coating the printed structures was arguably the riskiest step. Washcoating, the deposition of a catalytic ceramic layer onto a substrate, has historically been tricky on additively manufactured surfaces because of roughness, wettability, and adhesion concerns. The researchers prepared a stable suspension of platinum supported on gamma-alumina, with a median particle size near 200 nanometres, and dipped each monolith twelve times, spin-drying between immersions before calcining at 500 degrees Celsius. Drop tests and ultrasonic exposure showed relative washcoat losses below 2.5 percent for nearly all geometries, evidence that the rough printed surface actually helps by mechanically interlocking the ceramic layer. Scanning electron microscopy revealed porous, interconnected washcoat morphologies that should aid gas diffusion, along with an axial thickness profile that was thickest near the monolith ends, likely a consequence of slurry depletion and redistribution during drying.

Pressure-drop screening on longer, uncoated 150 millimetre samples showed that specific surface area alone does not dictate flow resistance. The Schwarz-diamond lattice with 5 millimetre cells, which packs the highest surface area of the set, also produced the highest pressure drop, while the open Schwarz-primitive structure with 10 millimetre cells flowed more freely than even the straight-channel hexagonal reference. In between, the gyroid geometries occupied a promising middle ground, their smooth continuous passages balancing gas-surface contact against hydrodynamic resistance.

The decisive test came on a three-cylinder 2.19-litre compressed natural gas engine operating at 1500 revolutions per minute and 150 newton-metres of load, with the air-fuel equivalence ratio swept from 0.9 to 1.4. Exhaust gas was routed through a bypass containing the washcoated monoliths, and inlet and outlet concentrations of methane, carbon monoxide, and nitrogen oxides were measured by Fourier-transform infrared spectroscopy and an exhaust gas analyser. Under stoichiometric conditions, the two best TPMS configurations exceeded 90 percent CO conversion, with the compact Schwarz-diamond SD(5,s) sample reaching approximately 99 percent. Normalising conversion by washcoat loading revealed that catalyst utilisation, not merely the amount of platinum deposited, differed substantially between topologies and was strongly influenced by unit cell orientation.

Equally striking was evidence of three-way catalytic behaviour. Methane and nitrogen oxide conversions peaked near stoichiometric operation, and under fuel-rich conditions the researchers observed ammonia forming inside the monolith, a signature of nitrogen oxides being reduced over platinum with CO and hydrogen acting as reductants. Because no external ammonia was introduced, its appearance confirmed that the printed carriers supported not just oxidation but reductive chemistry, extending their functional repertoire beyond simple CO cleanup. The team is careful to note that these results rest on one engine-tested sample per geometry, so they represent comparative trends within a proof of concept rather than statistically validated rankings.

When conversion and pressure drop were mapped together, the gyroid structure with 10 millimetre cells emerged as the standout, delivering higher washcoat-normalised CO conversion than the straight-channel reference while imposing lower pressure drop per unit length. The compact 5 millimetre designs bought their exceptional conversion with a clear backpressure penalty, whereas the most open Schwarz-primitive structures flowed beautifully but underused their catalyst. The authors conclude that performance is governed by the interplay of specific surface area, topology, effective tortuosity, flow distribution, and catalyst accessibility, and they propose TPMS architectures as a tunable design platform for application-specific catalytic supports. Before such printed converters reach production vehicles, they caution, long-term durability, thermal ageing, vibration resistance, benchmarking against commercial substrates, and techno-economic assessment will all need to be demonstrated. But the core message stands: a catalyst carrier generated from pure mathematics, printed layer by layer in steel, survived an engine’s exhaust and cleaned it remarkably well.

Subject of Research: Additively manufactured washcoated TPMS structures as catalytic converter carriers tested under real engine exhaust conditions

Article Title: Washcoated additively manufactured TPMS structures as catalytic converters for engines: A proof of concept

Article References: Mehdipour, F., Heinrich, J. E., Yu, Z., Kutscherauer, M., Tucker, M. R., Rubin, M., Wagner, U., Koch, T., & Dittmeyer, R. (2026). Washcoated additively manufactured TPMS structures as catalytic converters for engines: A proof of concept. Results in Engineering, 32, Article 113039. https://doi.org/10.1016/j.rineng.2026.113039

Image Credits: AI Generated

DOI: 10.1016/j.rineng.2026.113039

Keywords: additive manufacturing, TPMS, catalytic converter, gyroid, laser powder bed fusion, washcoating, CNG engine, exhaust aftertreatment, pressure drop, platinum catalyst, three-way catalysis, 316L stainless steel