When the Substrate Fights Back: Discovering a New Pore Formation Mechanism in Aluminium Laser Powder Bed Fusion

How synchrotron X-rays revealed that the substrate you print on can create defects you never expected — and what to do about it

Published: July 2026

Source: Additive Manufacturing, Volume 124 (2026), Article 105229

Keywords: LPBF, synchrotron X-ray imaging, intermetallic-induced pores, substrate microstructure, PA1 alloy, process map

For two decades, AlSi10Mg has reigned as the workhorse aluminium alloy for laser powder bed fusion (LPBF). Its near-eutectic composition, good castability, and thermal properties make it forgiving to print. But it has limits: tensile yield strength capped below 300 MPa, and a maximum service temperature of around 200 °C. For aerospace and autosport applications demanding both higher strength and thermal stability, a new generation of aluminium alloys is needed.

Enter PA1 — a custom-designed near-eutectic Al–Ni–Ce–Mn–Fe alloy engineered specifically for additive manufacturing. Its design philosophy is elegant: narrow the freezing range to just 2.8 °C to suppress hot cracking, and add transition metals and rare earth elements with low diffusivity and solubility to promote fine, thermally stable intermetallic reinforcement. Under the extreme cooling rates of LPBF (~10⁶ K/s), PA1 forms an extraordinarily fine cellular-eutectic microstructure with features smaller than 200 nanometres — two orders of magnitude finer than what conventional casting produces.

But here lies the catch. When you print PA1 powder onto an as-cast PA1 substrate — a common scenario in hybrid manufacturing or component repair — the substrate contains large columnar intermetallic phases (Al₉M₂, where M = Fe, Mn, Ni) that can reach hundreds of micrometres in length. These phases, harmless under the slow cooling of casting or the large melt pools of directed energy deposition, turn out to be anything but harmless under LPBF conditions.

A research collaboration spanning UCL, Brunel University, ESRF, IMDEA, PSI, and Oxford Brookes University deployed the Quad-ISOPR — a physical twin of Renishaw's RenAM 500Q industrial LPBF system — at the ESRF synchrotron's ID19 beamline. Using high-speed X-ray imaging at 40,000 frames per second with a pixel resolution of 4.3 μm, they watched, in real time, what happens when a laser melt pool traverses a substrate loaded with large intermetallic needles.

What they observed had never been documented before in LPBF. When the keyhole — the deep, narrow vapour cavity formed by intense laser vaporisation — passed over or adjacent to a large intermetallic particle, a localised instability occurred. The keyhole boundary touched the intermetallic surface, creating a gas–solid interface with reduced surface tension compared to the normal gas–liquid interface. Part of the keyhole then adhered to the intermetallic, detaching to form a pore. The researchers termed this entirely new defect an intermetallic-induced pore, or IM-pore.

Beyond the IM-pore discovery, the study systematically classified all defects observed during LPBF of PA1 on as-cast substrates into four categories:

• Hydrogen pores (H-pores). Small (<10 μm radius), spherical pores formed at the solidification front due to the 19× drop in hydrogen solubility between liquid and solid aluminium. While individually benign, they can coalesce into larger pores (~42 μm equivalent radius) that become mechanically detrimental.

• Keyhole pores (K-pores). Formed by keyhole instability and tip collapse, driven by Plateau-Rayleigh instability. Under unstable keyhole conditions (Es > 65 Js½/mm), both the number and total area of K-pores double.

• End-of-track defects (ETD). Two variants: end-of-track pores (keyhole collapse after laser shut-off in high-energy conditions) and end-of-track depressions (insufficient liquid feeding before solidification in moderate-energy conditions).

• Intermetallic-induced pores (IM-pores). The newly discovered mechanism — exclusive to substrates containing large intermetallic phases. Observed under all printing conditions except conduction mode, where the laser-beam/substrate interaction is minimal.

To isolate the substrate effect, the team performed identical LPBF experiments on two substrates: as-cast PA1 (coarse intermetallics) and as-rolled Al6061 (fine, sub-5 μm intermetallics). The difference was stark.

On the PA1 substrate, IM-pores formed consistently across all keyhole and depression-zone melting conditions. The large intermetallics also exacerbated keyhole instability: the number and total area of K-pores were significantly higher on PA1 than on Al6061 under comparable energy input. The spatial correlation between coarse intermetallics and pore initiation sites leaves little doubt — substrate microstructural heterogeneity directly drives defect formation during the critical first layers.

On the fine-grained Al6061 substrate, by contrast, no IM-pores were observed. The intermetallic size was simply too small relative to the melt pool dimensions to trigger the keyhole-attachment mechanism.

The study's process map delivers actionable guidance. For defect-minimised printing of PA1, the recommended regime is stable keyhole or depression zone melting at moderate specific energies (37.5–51 Js½/mm). Within this window, keyhole porosity is avoided, hydrogen pore populations are minimised, and end-of-track defects are manageable.

Critically, multi-layer experiments showed that the substrate microstructure effect diminishes with build height. After approximately 10 layers (~600 μm), heat accumulation and remelting homogenise the microstructure, and the intermetallic phases that do form are sub-micrometre in size — far less likely to trigger IM-pores. However, the phenomenon remains highly relevant for:

• Hybrid additive manufacturing, where AM is performed on conventionally cast or forged substrates containing coarse microstructures.

• Component repair, where laser cladding or LPBF is applied to existing parts with unknown substrate microstructures.

• Alloy systems prone to intermetallic formation, such as Al–Fe alloys, where needle-shaped intermetallics can form throughout the build under AM conditions, not just at the substrate interface.

• Dissimilar alloy printing, where the interface between two different materials is a natural site for intermetallic formation.

The IM-pore discovery adds an important new dimension to our understanding of LPBF defect formation. Until now, the LPBF community has focused primarily on process parameters — laser power, scan speed, hatch spacing — as the primary levers for defect control. This study demonstrates that the substrate is not merely a passive platform: its microstructure actively participates in, and can disrupt, the melting process.

For industry, the message is clear. When printing on non-AM substrates — particularly cast alloys with coarse intermetallic phases — consider pre-processing the substrate via rolling, heat treatment, or surface remelting to homogenise the microstructure before building. For alloy designers, the IM-pore mechanism should inform the selection of substrate materials and the design of initial layer strategies. And for the broader AM community, this work serves as a reminder that the physics of LPBF extends beyond the laser–powder interaction to encompass the entire thermal and metallurgical history of the build.

Reference: Guo, D., Zhu, C., Bhagavath, S., et al. (2026). Influence of substrate microstructure on defect formation during Al alloy laser powder bed fusion. Additive Manufacturing, 124, 105229.