Theory 22/Manufacturing & Metrology /Additive manufacturing/3 min
Porosity Comes From the Pool, Not the Powder
Quality control blames the powder. The defect may be born in the puddle.
Porosity in metal additive manufacturing is generated by oscillatory instabilities in the melt pool rather than by powder quality or energy density, so defect density should track pool oscillation frequency and be controllable by modulation rather than by tighter powder specifications.
01The evidence
- Melt pool dynamics and pore formation in selective laser melting have been characterised mechanistically, with pore formation tied to pool behaviour rather than to feedstock variability alone.Materials and Design
- Mixing-driven defects and composition evolution in multi-material additive manufacturing show that pool-level transport, not powder quality, can dominate defect generation.Advanced Science
02The tension
Powder quality and energy density are the standard levers, and both matter. But defect density varies between machines and between builds with the same powder and the same nominal parameters, and the variation tracks the pool's behaviour rather than the feedstock. That residual is usually treated as machine-specific noise.
03The theory
A laser melt pool is a driven fluid with a free surface. Under certain conditions it oscillates, and the oscillation can pinch off and trap vapour or eject material, leaving a pore. The instability is a property of the pool, not of the powder.
If porosity is instability-driven, then defect density should correlate with the oscillation frequency and amplitude rather than with powder size distribution, and it should be suppressible by modulating the laser at a frequency that damps the oscillation.
The prediction is that a modulated build should have lower porosity at the same energy density, and that in-situ monitoring of pool emission should predict porosity before the layer is complete.
This changes the economics of additive quality: powder specification is expensive, while modulation is a firmware change, and if the theory is right the cheaper lever is also the effective one.
04What this is not
Those models attribute porosity to feedstock or to insufficient energy. This attributes it to pool instability, which predicts that identical powder can give different porosity and that modulation can fix it.
05Predictions
- P1Defect density should correlate with pool oscillation metrics better than with powder size distribution.
- P2Laser modulation at a damping frequency should reduce porosity at constant energy density.
- P3In-situ pool monitoring should predict porosity before layer completion.
06What would refute it
Porosity tracking powder quality at controlled pool dynamics would restore the feedstock account.
07What it would change
It would replace powder specification with process modulation, and make in-situ monitoring a quality record.
08Evidence base
- Melt pool dynamics and pore formation in selective laser melting — Materials and Designhttps://www.sciencedirect.com/science/article/pii/S0264127525004630
- Mixing-driven defects in multi-material metal additive manufacturing — Advanced Sciencehttps://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.75723