Powder Qualification for Metal AM: What the Metrics Actually Mean

Powder is not a commodity input in metal AM. The same alloy from two different suppliers — or even two batches from the same supplier — can produce meaningfully different density, surface finish, and mechanical properties on an identical machine with identical parameters. Understanding what the characterisation metrics actually tell you (and what they do not) is the difference between a robust incoming quality procedure and a box-ticking exercise.

This article covers the metrics used in practice for LPBF, EBM, and binder jetting powder qualification, with particular attention to Ti-6Al-4V, AlSi10Mg, and IN718 — the three alloys where powder variability causes the most field problems.

Why Powder Matters: Degradation Mechanisms in Service

Virgin powder that passes all incoming tests can still fail in service. The key degradation modes are:

Satellite formation. During gas atomisation, small secondary droplets attach to larger particles before solidification. Satellites increase surface roughness, reduce apparent density, and worsen flowability. Satellite fraction increases with reuse cycles as particle impacts cause further agglomeration.

Oxidation. All metal powders oxidise on exposure to air or moisture. For titanium alloys, the oxygen pickup is permanent — oxides do not dissolve back into the matrix during melting at LPBF processing temperatures. Oxygen above ~0.25 wt% in Ti-6Al-4V reduces ductility below the ASTM F2924 requirement. Aluminium alloys form a stable Al₂O₃ skin on every particle; excessive oxide content manifests as oxide stringers in the as-built microstructure (Aboulkhair et al., 2019, Progress in Materials Science, doi:10.1016/j.pmatsci.2018.10.035).

Moisture absorption. Hygroscopic powders (particularly aluminium and copper alloys) absorb moisture in humid storage. During melting, the moisture releases hydrogen that forms gas porosity. Vacuum or inert-gas drying at 60–80°C for 4 hours before use is standard practice for Al alloys.

Particle size coarsening. Repeated thermal cycling in the powder bed sinters fine particles together, shifting the PSD toward larger sizes and wider distributions. Sieving after every build cycle is recommended; replacing powder with a high proportion of particles >63 µm is common practice.

The Powder Characterisation Tracker logs all these metrics per batch and flags out-of-specification values.

Particle Size Distribution: D10, D50, D90, and Span

PSD is measured by laser diffraction (ISO 13320) or dynamic image analysis (ISO 13322-2). The three key percentile values tell different things:

• D10: the particle size below which 10% of particles by volume fall. D10 governs the minimum feature resolution and filter blinding risk. For LPBF, D10 should be > 15 µm to avoid particles passing through the recoater mesh.
• D50: the median particle size. This is the headline specification number. Typical LPBF range: 20–45 µm. EBM typically runs coarser: 45–106 µm. Binder jetting is process-specific but generally 15–63 µm for metals.
• D90: the size below which 90% of particles fall. D90 drives the maximum layer thickness feasible without large particles causing surface roughness spikes or recoater interference. D90 should be ≤ 2× the planned layer thickness.

Span = (D90 − D10) / D50. A narrow span (< 1.5) indicates a tight, consistent distribution that packs predictably and flows uniformly. Wide-span distributions may have adequate D50 but contain enough fines (contributing to flow problems) or coarse particles (contributing to surface defects) to cause process issues.

For IN718, the AMS 7009 powder specification requires D10 ≥ 15 µm, D50 = 20–45 µm, and D90 ≤ 63 µm. Deviations outside this range are a direct hold point — do not use without engineering disposition.

Flow Rate: Hall Flow vs Carney Flow

Flowability governs how reliably powder spreads into a consistent, uniform layer. Two test methods are in common use:

Hall flow (ASTM B213 / ISO 4490): measures the time in seconds for 50 g of powder to flow through a 2.54 mm orifice funnel under gravity. Results are reported in s/50g. Lower is better (faster flow). Typical acceptance criteria:

Carney flow (ASTM B964): uses a 5.0 mm orifice — designed for powders that don't flow through the Hall funnel at all. Relevant for cohesive powders, highly spherical powders with poor inter-particle friction, and most aluminium alloys. If a powder does not flow through the Carney funnel either, the No Flow result is reported, which is typically a reject condition.

The limitation of both methods: they measure bulk flow under gravity, not the dynamic spreading behaviour under a recoater blade. Powders that pass Hall flow can still spread poorly on an actual machine if particle morphology is irregular. For critical qualification, complement flow testing with a spread density test (measuring the powder layer density in the actual build chamber under realistic conditions) and, ideally, an actual build test.

The Powder Characterisation Tracker accepts both Hall and Carney flow inputs and flags values outside process-specific acceptance windows.

Apparent and Tap Density: Hausner Ratio

Apparent density (ASTM B212 / ISO 3923-1): the density of powder poured loosely into a graduated cylinder, measured in g/cm³. It reflects how much air is trapped between particles in an undisturbed bed.

Tap density (ASTM B527 / ISO 3953): the density after a standardised number of taps (typically 3000 taps under ASTM B527), which settles and compacts the powder.

Hausner ratio = Tap density / Apparent density. This dimensionless ratio quantifies the inter-particle friction and cohesion:

A Hausner ratio > 1.25 is a reliable predictor of layer spreading problems, particularly in bidirectional recoater machines where the powder must flow in both directions. For reference, spherical gas-atomised Ti-6Al-4V in the 20–45 µm range typically achieves Hausner < 1.15; AlSi10Mg, with its irregular surface morphology from gas atomisation, is often in the 1.20–1.30 range.

Oxygen and Moisture Limits

Oxygen content is measured by inert gas fusion (ASTM E1019). The limits differ by material class:

For titanium, oxygen is the most critical single powder metric. Every reuse cycle in an oxygen-containing atmosphere (even ppm-level leaks in the process chamber) adds to cumulative powder oxygen. Many operators track a cumulative oxygen budget across reuse cycles and retire powder when the oxygen budget is exhausted, rather than waiting for a single-point test to exceed the limit.

For aluminium alloys, moisture (measured as loss on drying at 100°C for 2 hours) is equally critical. Aluminium powder with > 0.05 wt% moisture will generate hydrogen porosity during LPBF even at optimised parameters.

Recommended Test Sequence per Incoming Batch

Not every test is required at every delivery. A risk-stratified approach:

Tier 1 — Every batch (no exceptions):

1. Certificate of conformity (CoC) review — alloy chemistry per spec
2. PSD (D10/D50/D90/span) — laser diffraction
3. Hall or Carney flow rate
4. Oxygen content by inert gas fusion

Tier 2 — First article and annual re-qualification: 5. Apparent density and tap density (Hausner ratio) 6. Particle morphology — SEM imaging (assess satellite fraction, elongation, surface condition) 7. Moisture content

Tier 3 — New supplier qualification or after abnormal storage: 8. Full chemistry per alloy specification (ICP-OES or X-ray fluorescence) 9. Helium pycnometry (true density — flags hollow particles which will not consolidate) 10. Build trial on production machine with acceptance test per process specification

Log all results against the batch number in the Powder Characterisation Tracker. Trend analysis over successive batches from the same supplier often reveals slow degradation before a single-point test would catch it.

Reuse Policy

Powder reuse is economically essential — virgin Ti-6Al-4V powder costs $80–150/kg, and typically only 10–30% is consumed per build. But every reuse cycle changes the powder:

• Fines fraction decreases (sintered or filtered out)
• D50 shifts slightly coarser
• Oxygen content increases (Ti, Al) or remains stable (stainless, nickel)
• Satellite fraction increases

A common practice is blending fresh powder with recycled powder at a controlled ratio (e.g., 30% fresh / 70% recycled per cycle). This maintains PSD and oxygen within specification limits while reducing material cost. The critical requirement is that the blend is characterised, not just assumed to be acceptable.

ASTM F3456 (Standard Guide for Powder Reuse in Powder Bed Fusion Additive Manufacturing) provides a framework for establishing and documenting a reuse policy. It is not prescriptive about specific limits but requires that the policy be validated against the qualification build test.

Further reading

• Powder Characterisation Tracker — log and track PSD, flow, density, and chemistry per batch
• Ti-6Al-4V material data — process compatibility, oxygen limits, and key ASTM references
• AlSi10Mg material data — aluminium-specific moisture and oxide considerations
• IN718 material data — nickel superalloy powder limits per AMS 7009