Metal Powder Characterization for Additive Manufacturing: PSD, Flowability, Reuse and Quality Control

Metal powder characterization determines whether a powder lot is suitable for a specified additive manufacturing process. Effective metal powder testing covers particle size distribution, particle morphology, flowability, density, chemical composition, contamination and batch history. ISO/ASTM 52907:2019 provides methods for characterizing these properties, while ISO/ASTM 52928:2024 addresses powder lifecycle management. Neither standard establishes universal acceptance limits for every alloy or machine. Powder qualification requires representative samples, comparable test methods and acceptance criteria tied to the intended process.

1. What Is Metal Powder Characterization for Additive Manufacturing?

Metal powder characterization is the measurement and evaluation of feedstock properties before powder enters an additive manufacturing process or returns to production after use. It helps you determine whether the material matches its specification and whether changes in powder condition could affect processing.

For metal powder for additive manufacturing, the evaluation covers more than particle size. Two lots of nominally identical 316L stainless steel powder can have similar D50 values but different particle shapes, surface conditions, flow times or oxygen contents. A matching alloy designation does not establish equivalent powder behavior.

ISO/ASTM 52907:2019 identifies the main properties and methods used for powder characterization, including chemical composition, particle size, morphology, characteristic densities, flowability and contamination. The standard also addresses documentation, sampling and traceability.

A test report states measured properties. Production qualification also requires limits defined for the alloy, process and application.

Supplier lots with the same nominal particle size range may behave differently during laser powder bed fusion, particularly after reuse.

When comparing metal 3D printing technologies, distinguish the feedstock requirements of PBF, DED and binder jetting before transferring powder limits between processes.

1.1 Powder Identity, Lot Selection and Representative Sampling

Before comparing particle size distribution or powder flowability, establish exactly which powder was tested.

A useful powder record identifies the alloy grade, powder production route, supplier lot, material condition and relevant processing history. For recovered material, it should also record screening, blending, build exposure and the relationship between the original and subsequent powder batches.

Representative sampling is equally important. Metal powder can segregate during handling, recovery and transfer. A sample collected from one location may not represent the particle size or contamination condition of the entire lot.

ISO/ASTM 52907:2019 addresses representative powder sampling and identifies ISO 3954 and ASTM B215 among the applicable sampling methods. The procedure must preserve sample identity and avoid introducing contamination.

The 316L case tracks a recovered batch separately from the original feedstock.

The original material was virgin powder, argon atomized 316L stainless steel, processed using PBF-LB/M. After two build cycles, 80.0 kg of screened recovered powder was blended with 20.0 kg of fresh powder. The mixture was handled in a closed environment and mixed for 20 minutes. Incremental samples were then collected from the upper, middle and lower regions to form a composite sample.

The subsequent test results describe the 80:20 blend after recovery, sieving and sampling, not unmixed recovered powder.

Every PSD, morphology, flow and chemistry result must refer to that blended condition.

2. Particle Size Distribution: D10, D50, D90 and Test Method

Particle size distribution (PSD) describes the proportion of particles in defined size classes. In a cumulative volume distribution, D10, D50 and D90 are the diameters below which 10%, 50% and 90% of the measured particle volume falls. D50 alone hides both the fine and coarse tails.

The supplied 316L case used ISO 13320:2020 laser diffraction with dry dispersion. Three independent samples were collected per powder condition, with three repeat measurements per sample. The original powder measured D10 21.0 μm, D50 34.5 μm and D90 49.0 μm. After recovery, 63 μm sieving and an 80:20 blend, the values were 22.0, 35.2 and 50.5 μm.

The small shift in median size does not erase the tail changes. The measured fraction below 20 μm decreased from 8.0% to 6.7%, while the fraction above 50 μm rose from 8.0% to 10.7%. These are case-specific laser diffraction volume fractions, not universal tolerance bands. The particles removed by screening, segregation and blending history all matter when interpreting the change.

2.1 Laser Diffraction, Imaging and Sieving Are Not Interchangeable

For particle size analysis, the chosen method changes the meaning of the result. Laser diffraction under ISO 13320 reports an optical equivalent diameter distribution, typically on a volume basis. Image analysis describes projected particle dimensions and, where specified, shape metrics. Sieving reports the mass retained or passed through an aperture. They do not measure identical quantities.

Method	Typical output	What it answers	Comparison limit
Laser diffraction, ISO 13320	Volume-based D10/D50/D90	Has the apparent size distribution shifted?	State dispersion, optical model and sample condition.
Static/dynamic image analysis, ISO 13322 series	Particle dimensions, counts and shape metrics	Are irregular or satellite-bearing particles present?	Define measured diameter, threshold and counted population.
Sieving, applicable sieve method	Mass retained/passed per mesh	Is oversized material removed?	Sieve aperture is not a D90 equivalent.

For incoming inspection, request the same test method and reporting basis across batches. Powder sieving may complement, rather than replace, fine-particle laser diffraction particle size analysis. Controlled metal additive manufacturing capabilities depend on stable, documented powder test methods across repeat builds.

3. Powder Morphology: Sphericity, Satellites and Agglomerates

Powder morphology describes particle shape, surface condition and features such as attached satellites, elongated grains and powder agglomeration. Gas atomized particles are often approximately spherical, but a powder labeled spherical still needs measurement or inspection when morphology is a quality requirement.

Particle shape analysis can reveal changes that a PSD curve does not show. The 316L case used ISO 13322-1:2014 static image analysis. Under the case definition of visibly agglomerated, satellite-bearing or non-spherical particles, 32 of 1,200 particles (2.67%) were classified as abnormal in the original powder. The blended reused condition had 54 of 1,200 (4.50%).

These percentages describe this case’s classification rule and observed particles. They are not material-independent sphericity criteria and should not be compared against another laboratory’s result unless its imaging, definitions and sampling are aligned. Unusual particle surfaces can contribute to friction and local spreading defects, but a micrograph cannot alone establish machine recoating performance.

4. Powder Flowability vs Spreadability in LPBF

Powder flowability describes how a bulk powder moves under a specified test condition. Spreadability describes how material is distributed into a layer by a blade or roller in a real build environment. They are related process concerns, not interchangeable test results.

The powder flowability testing in the 316L case used ISO 4490:2018 Hall flow, 50.0 g per run, at 23 ± 2 °C and 45 ± 10% relative humidity. Original powder readings were 16.2, 16.4 and 16.3 s/50 g, averaging 16.30. The 80:20 reused blend returned 17.2, 17.5 and 17.4 s/50 g, averaging 17.37. The latter was slower by 1.07 s/50 g, approximately 6.5% relative to the original mean.

Both groups remained below the case acceptance value of 20 s/50 g. The result alone says nothing conclusive about the local powder-bed packing density, blade streaking or thin-layer coverage. Powder flow properties may also change with humidity, particle friction, cohesion, test apparatus and conditioning.

4.1 Hall/Carney Flow Results Cannot Alone Qualify Recoating

A Hall flowmeter measurement reports how long a defined mass takes to pass through its calibrated funnel. Carney methods use a different funnel and test mass. Never compare a Hall s/50 g time directly with a Carney time as though the numbers were on one scale.

In the case build, the same recoater setup was used with a 40 μm layer thickness. Both material conditions completed ten consecutive layer checks. The reported observation for the blended powder was continuous spreading without visible cross-track streaks, exposed patches or blade drag. Ten observed layers support that limited observation; they do not constitute a complete spreadability or production qualification campaign.

A better evaluation links a standardized flow measurement with direct layer observations and the actual application requirements. Recoater type, blade speed, layer thickness and surface condition can change the answer. additive manufacturing DFM considerations also affect where layer defects turn into problematic geometry or surfaces.

5. Apparent, Tapped and Skeletal Density

A powder density value is incomplete unless the density type and test method are named. Apparent density describes powder filling a container under a prescribed loose-fill procedure. Tapped density measures volume after standardized tapping. Skeletal density measures the particle material volume by a different principle and cannot be substituted for either bulk measurement.

The case used ISO 3923-1:2018 for apparent density and ISO 3953:2011 for tapped density. Mean apparent density changed from 4.31 to 4.28 g/cm³. Mean tapped density changed from 4.86 to 4.83 g/cm³. The reported values show a slight decline in both loose and tapped packing measures after the reuse and blending sequence.

Neither number is the same as a laser powder-bed packing density measured in situ, and neither determines the built part’s final relative density. The Hausner ratio, calculated as tapped density divided by apparent density, is approximately 1.128 for the original powder and 1.129 for the blend. It is useful as an index of bulk powder packing response, not an independent approval of recoating.

Terminology and alloy state matter beyond the powder itself. The metal material properties and specifications reference provides context for distinguishing feedstock characteristics from properties assigned to manufactured material.

6. Chemistry, Oxygen Pickup and Contamination Control

Metal powder chemistry can change through exposure during build cycles, recovery, transfer, storage and blending. Oxygen, nitrogen and hydrogen are important to control where the selected alloy and qualified process make them critical. Metal powder contamination can also come from foreign particles, other alloys, condensate, handling equipment or moisture.

The 316L case reported oxygen increasing from 0.018 to 0.024 wt.% and nitrogen from 0.055 to 0.060 wt.%. Hydrogen changed from 0.0003 to 0.0004 wt.%. These are measured values for the two reported powder states; they do not establish a universal rate of oxygen pickup per build cycle. The case used ASTM E1019-24 for these elemental determinations.

The case also recorded 105 °C mass loss of 0.010 and 0.013 wt.% for the original and blended powder. This loss-on-heating test is not a specific water assay, so its result must not be relabeled as actual moisture content. Screening for foreign metallic particles reported no distinguishable other-alloy particles in the examined material.

When writing a purchasing specification, distinguish material limits from how you obtain the numbers. ISO/ASTM 52907 describes characterization methods and documentation; it does not set common oxygen or nitrogen acceptance limits for 316L, Ti-6Al-4V and Inconel 718. The material specification and qualified process define the relevant chemistry limits. For safety-sensitive applications, medical metal additive manufacturing quality makes that separation particularly important.

7. Reused vs Virgin Metal Powder: Requalification and Reblend Control

Powder reuse is a controlled material state, not a synonym for unrestricted recycling. A recycled powder may be sieved, blended, sampled and tested before it is returned to a machine. The original supplier certificate cannot by itself establish the condition of material that has experienced subsequent builds.

In the submitted 316L case, recovered powder was sieved through a nominal 63 μm aperture. An 80.0 kg recovered fraction was combined with 20.0 kg fresh powder to give 100.0 kg of an 80:20 blend. The blended sample was checked for PSD, shape, Hall flow, apparent density, tapped density, oxygen, nitrogen, hydrogen and foreign particles.

The resulting comparison shows why no single metric should carry the decision. The blend had a slightly higher D90, longer Hall flow time, more classified abnormal particles and higher oxygen. Yet the reported parameters remained within the case acceptance limits. That supports progression to the next defined process validation stage, not unconditional approval of all future builds.

316L powder characteristic	Original powder	80:20 reused blend	Interpretation for this case
D10 / D50 / D90, μm	21.0 / 34.5 / 49.0	22.0 / 35.2 / 50.5	Coarse tail shifted; D90 remains within 55 μm case limit.
Abnormal morphology, count	32/1200 (2.67%)	54/1200 (4.50%)	Both below the 5.0% case threshold.
Hall flow, mean s/50 g	16.30	17.37	Slower, but both below 20 s/50 g.
Apparent density, g/cm³	4.31	4.28	Both above the 4.10 g/cm³ case threshold.
Tapped density, g/cm³	4.86	4.83	Small reduction; no separate tapped-density limit given.
Oxygen, wt.%	0.018	0.024	Both below the 0.030 wt.% case threshold.
Nitrogen, wt.%	0.055	0.060	Both below the 0.080 wt.% case threshold.
105 °C mass loss, wt.%	0.010	0.013	Both below the 0.020 wt.% case threshold.

For a supplier or contract manufacturer planning repeatable low volume manufacturing, these differences become meaningful when they are tracked consistently against the same process and test definitions. Switching the sampling method or powder blending rule midway through the comparison breaks that continuity.

7.1 Reuse Metrics, Sieving, Blending and Retirement Rules

ISO/ASTM 52928:2024 addresses powder lifecycle management and defines a controlled approach to metal powder recycling and reuse. The practical reuse metric may involve build count, exposure duration or energy, powder test results and qualification of the relevant process. It is not automatically equivalent to the number of times powder has passed through a sieve.

A fixed answer such as “reuse 316L ten times” ignores thermal exposure, machine operation, powder bed conditions and replenishment strategy. For the submitted case, the defined history is two completed build cycles plus a specified 80:20 blend. Those facts describe this lot only. They do not set a maximum reuse limit for other batches.

Operational records should connect recovery, screening, blending, sampling, testing and either release or retirement to an identifiable powder population. Material that no longer meets the relevant specification should not be returned to the same qualified production use just because it remains free flowing.

8. Lot Traceability and Powder Certificate of Analysis

Lot traceability is more than retaining a powder label. It allows a team to reconstruct how the material moved from receipt through storage, splitting, builds, recovery, sieving and mixing. ISO/ASTM 52928 describes the need to track powder lifecycle events; ISO/ASTM 52907 connects powder characterization to identity and documentation.

A material certificate or powder certificate of analysis (CoA) should identify which lot and material condition the results cover. At minimum, the purchasing team needs alloy identity, production route, applicable specification, sampling status, test method, result units, analysis date and traceable lot reference. Used powder adds a documented reprocessing and blending history.

In the case, the initial batch and the recovered 80:20 mixture are distinguishable states. Linking a D90 of 50.5 μm or an oxygen result of 0.024 wt.% to only the original batch label would misrepresent the sample. A practical event log should show material movement and quantity as well as the changed state after blending.

9. Metal Powder Testing and Incoming Acceptance Checklist

Incoming powder acceptance is a powder quality control decision for a specified use. Metal powder testing should therefore match measurable attributes to stated limits, test methods and the actual build route. A suitable checklist begins with identity and representative sampling, then compares the current powder state with the agreed acceptance requirements.

Incoming inspection field	Information to specify or record	Why it changes the decision
Material and lot	Alloy grade, atomization route, lot ID, virgin or blended status	Prevents one powder population being mistaken for another.
Sampling	Method, location, subdivision and condition	Connects test outcomes to a representative population.
PSD	Test standard, distribution basis, D10/D50/D90 and relevant tails	Allows meaningful comparison and oversized control.
Morphology	Imaging method, number of particles, classification criteria	Makes particle-shape statistics interpretable.
Flow and density	Apparatus, conditioning, units, loose/tapped density type	Avoids mixing unlike methods and properties.
Chemistry and contamination	Element methods, applicable limits, handling status	Detects changes relevant to material requirements.
Reuse and release	Build exposure, sieving, blend proportions, reuse metric, disposition	Maintains control over returned feedstock.

For the reported 316L blend, the project-specific checks included D10 at least 18 μm, D90 no more than 55 μm, Hall flow no more than 20 s/50 g, apparent density at least 4.10 g/cm³, oxygen no more than 0.030 wt.%, nitrogen no more than 0.080 wt.%, abnormal particle count no more than 5.0%, and 105 °C mass loss no more than 0.020 wt.%. These are the supplied case acceptance limits, not requirements imposed by ISO/ASTM 52907 or a universal 316L feedstock specification.

The blend met the listed numerical limits, and the ten-layer observation met the stated spreading condition. Its recorded disposition was to proceed to the next stage of build-process validation. Final part acceptance requires additional evidence appropriate to the application, potentially including density, discontinuities, dimensions and mechanical performance. Critical bore and datum requirements may require CNC machining for metal AM critical features and a separate inspection plan aligned with GD&T and tolerance standards.

9.1 What ISO/ASTM 52907 and 52928 Cover and Do Not Specify

ISO/ASTM 52907:2019 is a powder characterization methods standard. It identifies relevant properties and accepted approaches to sampling, measurement, reporting and documentation. ISO/ASTM 52928:2024 addresses powder lifecycle management, including recovered material and controls for reuse. Neither document supplies a universal D90, Hall flow time, oxygen ceiling or maximum reuse count that qualifies every metal powder for every machine.

An RFQ can be specific without exposing confidential production information. Define the alloy, selected process, material condition, required measurement methods, agreed limits, reusable-powder controls and reporting fields. A purchaser comparing prototype build routes can also reference metal 3D printing prototype evaluation without presenting a powder test result as proof of finished-part capability.

10. Handling, Storage and EHS Interface

Metal powder handling affects both product quality and metal powder safety. Uncontrolled storage and transfer can change powder moisture content or introduce foreign material. Fine metallic powders may also present inhalation, fire or explosion hazards depending on alloy, particle condition and the handling environment.

ISO/ASTM 52931:2023 addresses environmental, health and safety principles for metallic materials in additive manufacturing. It is not a replacement for powder chemistry limits or process acceptance requirements. Powder containers, transfer equipment, storage atmosphere, cleaning controls and waste arrangements should be selected using the material-specific safety assessment and applicable facility controls.

The 316L case described sealed argon storage and dry transfer with reduced air exposure. Those conditions give context to the reported powder history, but should not be generalized into a complete safety program or proof that all stored powder is production ready.

11. Questions Engineers Ask Before Releasing a Powder Lot

11.1 What Powder PSD Is Suitable for LPBF?

There is no single suitable particle size distribution for every LPBF machine, alloy and layer thickness. The permissible fines, coarse tail and D-values depend on the qualified feedstock and recoating process. Specify the measurement method before comparing a nominal size range. The case blend had D10 22.0 μm and D90 50.5 μm against its own documented limits; that does not establish a recommended distribution for other systems.

11.2 How Many Times Can Metal 3D Printing Powder Be Reused?

No universal number applies. Powder reuse should be controlled through a validated powder lifecycle management strategy and acceptance criteria tied to measured powder changes and machine exposure. The case involved two build cycles and an 80:20 recovered-to-fresh blend, not a qualified maximum cycle count.

11.3 Does Good Powder Flowability Guarantee a Uniform Layer?

No. A Hall flow test characterizes bulk gravity flow under its own procedure, whereas powder spreadability depends on the recoater, layer thickness, powder condition and process settings. The 316L blend met its Hall flow limit and completed ten observed recoating layers, but a longer production run and final-part checks require separate validation.

11.4 What Tests and Records Should a Metal Powder CoA Include?

A useful powder CoA identifies alloy and lot, material condition, sampling method, tested properties, methods, units and the relevant specification. For reusable feedstock, add recovery, screening, blending, exposure and current release status. The purchaser should agree which tests apply and which limits determine acceptance rather than treating every possible powder characterization result as mandatory.

The next production decision should retain the lot identity, applicable limits, test methods and powder history alongside the results.