Corrosion resistance of additively manufactured aluminium alloys for marine applications

Additively manufactured (AM) aluminium alloys enable tailored microstructures and superior mechanical properties compared to cast counterparts due to rapid melting/solidification and fine grains. AM AlSi10Mg has been widely studied and often used as a benchmark for AM Al-alloys, showing improved mechanical integrity and better corrosion resistance than cast Al–Si–Mg alloys because of reduced Volta potential differences of Si relative to Al. However, AM AlSi10Mg is susceptible to localized corrosion due to cathodic Si particles. For marine applications, there is a need for alloys with even higher corrosion resistance. Metastable Al alloys produced via non-equilibrium processing can provide enhanced corrosion resistance through homogeneous element distribution and supersaturated solid solutions. Mn and Mg commonly strengthen Al; Mn also improves general corrosion resistance due to potentials close to the Al matrix. Additional elements such as Cr and Zr can further influence corrosion and microstructure (e.g., Cr incorporation in passive films increasing pitting potential; Zr forming Al3Zr dispersoids that refine detrimental intermetallics). Heat treatments can change particle distributions and affect localized corrosion. This study investigates localized corrosion susceptibility and mechanisms in a newly designed AM Al–Mn–Cr–Zr alloy (Alloy C) versus AM AlSi10Mg, assessing microstructure, Volta potential differences, corrosion initiation/evolution, and electrochemical performance in NaCl and natural seawater for as-printed (AP) and heat-treated (HT) states.

Prior work established AlSi10Mg as a benchmark AM alloy with finer microstructures, improved mechanical properties, and generally better corrosion resistance than cast alloys, though cathodic Si particles drive localized corrosion. Studies have shown Si particle size and distribution influence Volta potential differences and corrosion driving force; higher ΔV at melt pool boundaries (MPB) correlates with more localized attack. Metastable Al alloys with extended solid solubility can exhibit excellent corrosion resistance. Mn additions up to ~5 wt% are feasible in AM and confer general corrosion resistance both in solution and as precipitates, due to potentials similar to the Al matrix. Cr, despite low solubility in Al, can be introduced via non-equilibrium techniques and has been linked to more protective passive films and higher pitting potentials. Zr additions form Al3Zr dispersoids, improve recrystallization resistance, and may enhance corrosion resistance by refining/intercepting detrimental intermetallics; however, heat treatments can alter particle distributions and influence localized corrosion behaviour. Localized corrosion mechanisms in AM alloys (e.g., AlSi10Mg, AA7075) have been linked to microstructural heterogeneity, particle chemistry, and ΔV measured by SKPFM.

Materials and powders: Pre-alloyed Al–Mn–Cr–Zr powder (nitrogen gas atomized; Höganäs AB) and standard AlSi10Mg powder (EOS GmbH). ICP-OES composition and particle size distributions provided (e.g., Alloy C ~Al–5.0Mn–0.8Cr–0.6Zr–0.17Fe–Si; D50 ≈ 29 μm; AlSi10Mg D50 ≈ 48 μm). Additive manufacturing and heat treatments: PBF-LB on EOS M290 (Yb-fiber laser 400 W, 370 W nominal). Parameters: power 370 W, scan speed 1300 mm/s, hatch 0.13 mm, scan rotation 67°/layer, layer thickness 0.03 mm; relative density >99.7%. Specimens ~105×10×12 mm. Heat treatments: Alloy C AP (as-printed, no heat treatment); Alloy C HT (350 °C, 24 h); AlSi10Mg (stress relief 250 °C, 2 h). Build plate pre-heating: 150 °C for Alloy C; 35 °C for AlSi10Mg. Contour strategy: no contours for Alloy C; AlSi10Mg with contour (320 W, 560 mm/s). Microstructural characterization: Cross-sections prepared parallel to build direction (XZ), mounted in conductive epoxy, ground/polished to 1 μm and OPS finish. Optical microscopy (ZEISS Axioscope 7) for stitched sections and porosity/density via cut-section analysis (ImageJ; detection limit ~1 μm). SEM (Zeiss Gemini 450 FE-SEM; 5 kV, ~1 nA) with Oxford ULTIM MAX EDS for sub-micron elemental mapping; additional SEM (Gemini 300, 10 kV) for post-exposure surfaces. Electrochemical heterogeneity mapping: SKPFM (Park Systems NX10, EFM mode) with ElectriMulti75-G probe (<25 nm tip, 75 kHz), 2 nm lift height, 5 V tip bias to obtain Volta potential maps and ΔV between particles and Al matrix. Exposure tests and image analysis: Polished samples exposed to 3.5% NaCl for 15 min, 1 h, and 24 h; natural seawater also used for electrochemical tests (pH ~7.86, salinity ~31.9 psu, Kristineberg Center). Post-immersion SEM to assess initiation and morphology. Quantification via ImageJ with Trainable Weka Segmentation and SEM Particle Segmentation plugins to obtain particle size distributions, cumulative cathode area, micro-galvanic couple interface length, and corroded area. Electrochemical testing: After 24 h immersion, OCP for 1 h followed by potentiodynamic polarization (PARSTAT 3000A-DX + Gamry ECM8; three-electrode cell with Ag/AgCl (sat. KCl) reference and Pt/graphite counter; scan from −0.25 V vs OCP to 0 V vs Ag/AgCl at 1 mV/s). Tafel extrapolation (±0.1 V from Ecorr) to determine icorr; duplicate samples measured. EIS: ±10 mV AC at OCP from 100 kHz to 10 mHz (fitting limited to 100 kHz–100 mHz due to low-frequency noise) analyzed with ZSimp Win using a Randles circuit with a constant phase element (CPE) to extract Rs, Y0, n, and charge transfer resistance (Rct).

Microstructure and particles: Both alloys showed high relative density (~99.8–99.9%). AlSi10Mg exhibited a fine eutectic Al–Si network with regional differences between bulk and border (indicative of different cooling rates). Alloy C (AP) contained dispersed nanometric Mn-rich precipitates: needle/rod-shaped (<100 nm) in melt pool centers and larger spherical/circular precipitates (200–400 nm) at MPBs; at borders, MPB precipitates reached ~1–2 μm. EDS indicated Mn and Cr enrichment at bulk MPB precipitates; border MPB precipitates were mainly Mn-rich. Heat treatment (HT) led to fine Al-(Mn,Cr) precipitates forming in the matrix and grain boundaries with slightly larger MPB precipitates, consistent with precipitation hardening. Volta potential (SKPFM) and particle statistics: In all alloys, second-phase particles were cathodic relative to the Al matrix. For AlSi10Mg, Si particles had higher ΔV at MPBs (median ~0.23 V) than in bulk (~0.18 V), consistent with greater localized corrosion along MPBs. Alloy C AP showed lowest median ΔV for precipitates within MPBs, higher in cold-worked/other regions, and highest in bulk (needle-like precipitates). Alloy C HT showed MPB precipitates with greater ΔV than bulk. Particle geometry strongly influenced susceptibility: AlSi10Mg had the greatest cumulative cathode area and micro-galvanic couple (MGC) interface length despite fewer discrete particles (Si network), whereas Alloy C AP had larger individual precipitates and Alloy C HT had a higher number density of smaller precipitates. Corrosion initiation and evolution (NaCl): Localized attack initiated in the Al matrix adjacent to cathodic particles. AlSi10Mg showed the most extensive early damage, following the eutectic network, driven by large cumulative cathodic area and MGC interface length. Alloy C showed localized attack around discrete precipitates; in Alloy C AP, higher ΔV needle/rod precipitates were preferentially attacked. In Alloy C HT, MPB particles were most susceptible due to higher ΔV. Printing strategy affected initiation: lack of contour in Alloy C produced border porosity/coarser MPB precipitates that acted as initiators; AlSi10Mg with contour showed less localized border initiation. After 24 h NaCl exposure, AlSi10Mg developed widespread corrosion products across the surface, while Alloy C AP/HT maintained localized but intensified damage; cumulative corroded area was higher in Alloy C AP than HT. Overall, ΔV drove initiation, while local particle density governed damage intensity over time. Potentiodynamic polarization (24 h exposure): In NaCl and seawater, Alloy C (AP, HT) exhibited passive regions up to higher pitting potentials than AlSi10Mg; AlSi10Mg in seawater remained active without passivation and had the highest Ecorr. Alloy C had significantly lower icorr and ipass than AlSi10Mg in both media. Representative values (from Tafel analysis): NaCl—AlSi10Mg icorr ≈ 2.95±0.54 μA/cm², Ecorr ≈ −0.73 V vs Ag/AgCl; Alloy C AP icorr ≈ 0.036±0.01 μA/cm² (≈ two orders of magnitude lower), Ecorr ≈ −0.71 V; Alloy C HT icorr ≈ 0.035±0.02 μA/cm², Ecorr ≈ −0.82 V. Seawater—AlSi10Mg icorr ≈ 2.07±0.29 μA/cm², Ecorr ≈ −0.60 V (no passive region); Alloy C AP icorr ≈ 0.31±0.03 μA/cm², Ecorr ≈ −0.98 V; Alloy C HT icorr ≈ 0.34±0.09 μA/cm², Ecorr ≈ −0.91 V. Heat treatment slightly reduced pitting resistance in NaCl (~0.1 V more negative Epit). EIS (24 h): In NaCl, total low-frequency impedance was ~3.7 kΩ for AlSi10Mg vs ~272 kΩ (Alloy C AP) and ~446 kΩ (Alloy C HT), indicating markedly higher resistance for Alloy C. In seawater, total impedance was ~13 kΩ (AlSi10Mg), ~56 kΩ (Alloy C AP), and ~44 kΩ (Alloy C HT). Nyquist semi-circles were larger for Alloy C, consistent with higher corrosion resistance. Phase angle behavior suggested denser, more protective passive films for Alloy C (lower capacitance, broader peaks, evidence of slower diffusion processes) than for AlSi10Mg. Overall performance: Alloy C (both AP and HT) outperformed AlSi10Mg in NaCl and natural seawater in terms of lower icorr, higher Epit, larger impedance, and more stable passivity. The coarser, discrete-precipitate microstructure of Alloy C favored formation of a more protective passive film compared to the fine eutectic Si network in AlSi10Mg.

Findings link microstructural heterogeneity and electrochemical contrast to localized corrosion behavior. In AlSi10Mg, a continuous eutectic Si network increases cumulative cathodic area and MGC interface length, promoting widespread, network-following attack even when ΔV is modest. In Alloy C, discrete Mn/Cr-containing precipitates produce localized micro-galvanic cells; initiation is dominated by sites with higher ΔV (e.g., needle/rod precipitates in AP, MPB precipitates in HT), but the overall extent is moderated by smaller cumulative cathodic area and interface length. Heat treatment refines precipitate distributions, generating many smaller particles with lower ΔV and reducing circumferential pitting likelihood, albeit with a slight decrease in pitting potential in NaCl, potentially due to grain-boundary precipitates. Electrochemical tests corroborate the microstructural interpretation: Alloy C forms denser, more stable passive films with extended passive regions and higher pitting resistance; AlSi10Mg lacks passivity in seawater and shows higher passive currents and lower impedance. Printing strategy (presence of a contour) affects border porosity and MPB precipitate morphology, influencing initiation hotspots. Overall, Alloy C’s chemistry (Mn, Cr, Zr) and precipitate characteristics support passive film resilience and superior corrosion resistance in chloride-bearing environments, making it more suitable for marine applications than AlSi10Mg.

Two AM Al alloys (AlSi10Mg and a newly designed Al–Mn–Cr–Zr Alloy C) were fabricated via PBF-LB, characterized, and evaluated in 3.5% NaCl and natural seawater. AlSi10Mg exhibits a fine eutectic Si network that promotes widespread localized corrosion due to large cumulative cathodic area and interface length, especially at MPBs. Alloy C presents discrete Mn/Cr-rich precipitates (coarser in AP; refined and more numerous after HT), with localized attack governed by ΔV and particle density. Electrochemical measurements demonstrate substantially lower corrosion and passive currents, higher pitting potentials, and one to two orders of magnitude higher impedance for Alloy C compared to AlSi10Mg, with AlSi10Mg failing to passivate in seawater. The coarser microstructure and precipitate chemistry of Alloy C favor formation of a stable, protective passive film and extended passivity. Heat treatment slightly reduces pitting potential in NaCl but maintains overall superior resistance; it also refines precipitates, reducing pitting likelihood. The results indicate that the newly designed AM Al–Mn–Cr–Zr alloy offers suitable and improved corrosion resistance for marine environments compared to AM AlSi10Mg. Future work could optimize printing strategies (e.g., contour use) and heat treatments to minimize initiation sites (porosity/MPB precipitates), and investigate long-term/severity conditions to further validate marine performance.

• Porosity/density estimation via cut-section optical analysis (ImageJ) is destructive and has a detection limit of ~1 μm; alternative non-destructive methods (CT, Archimedes) were not used here.
• EIS fitting was limited to 100 kHz–100 mHz due to low-frequency noise, which may affect resolution of very slow interfacial processes.
• Differences in printing strategy (use of contour for AlSi10Mg vs none for Alloy C) influenced border porosity/MPB precipitates and corrosion initiation, introducing a processing variable alongside alloy chemistry.
• Corrosion initiation/evolution was assessed at relatively short exposures (15 min, 1 h, 24 h), focusing on early-stage behaviour.
• Surfaces examined were parallel to the build direction; anisotropy effects for other orientations were not explored.