For many applications, AM is not economically viable without powder re-use. Powder bed fusion and binder jetting processes both have low ‘per pass’ incorporation rates and simply disposing of the residual powder is not an option, especially when using expensive metal powder feedstocks.
However, every pass through the printer can affect both chemical and physical properties of the powder. Physical changes that affect critical behaviours such as flowability and spreadability can be challenging to assess when establishing a robust re-use strategy.
Changed forever? The effect of passage through the printer
Particle morphology – size and shape – is widely recognised as performance-defining for AM powders and it can change markedly due to passage through the printer.
Metal powders for AM are typically less than 50µm in size, though coarser particles, in the size range 50 to 150µm, are better suited to laser melt deposition or electron beam melting. Particle size distribution is controlled during manufacture and by post-processing steps such as scalping. Passage through a powder bed printer tends to reduce the level of fines, particles around 10 – 20µm, via processes such as agglomeration and melting. Spattering, the splashing of molten metal can produce relatively large particles but larger particles from any source may be swept from the build platform by the recoater arm and not used. A narrowing of particle size distribution is typically the net result of all these effects.
When it comes to shape, manufacturing method is highly influential and virgin particles may exhibit irregularity and/or satellites, smaller particle grains adhering to the surface of larger particles. Temperature/melting can deform particle shape while the stresses applied during processing may remove satellites, potentially impacting both shape and size.
Morphological changes may directly affect performance in the printer. For example, changes in particle size alter response to heat, the ease and speed with which a particle melts, but equally importantly they influence process efficiency and print quality by changing characteristics such as packing performance and flowability. These relationships are harder to identify, but crucial.
Quantifying flowability, a comparison of used and virgin powders
LPW Technology, now part of Carpenter Additive, a market leading manufacturer of metal powders for AM, carried out a study of the flowability of re-used powders to support a multinational client establishing re-use strategies1. Basic Flowability Energy (BFE), a dynamic powder property measured using the FT4 Powder Rheometer, was used to quantify flowability. The aim was not just to detect differences – between virgin and used powder – but also to assess options for returning used powder to a comparable state, should it have deteriorated.
LPW invested in the capability to measure dynamic powder properties when it became apparent that traditional techniques – angle of repose, tapped density and Hall Flow Index, all carried out to ASTM standards – could not reliably rationalise observed differences in AM performance. Powder batches with ‘identical’ specifications were found to print inconsistently, particularly in newer, high speed machines. Furthermore, some traditional tests proved unsuitable for certain powders. For example, Hall flow rate cannot differentiate less free-flowing powders, those that produce a null result. Dynamic testing is suitable for all powders and has proven both relevant and highly differentiating for AM powders2.
The BFE values of virgin and used samples, and blends of the two, are shown above. The virgin powder has the lowest BFE and the used powder the highest. Passage through the printer has significantly altered flowability, potentially resulting in inferior print performance. Sieving has a minor impact on flowability but does not re-establish the state of the virgin material.
Blending is an obvious strategy for re-use so blends with different ratios of used and virgin materials were also tested. Perhaps unsurprisingly, the blend containing the highest proportion of virgin material produces a BFE value closest to the unused material. However, the 50:50 blend generated a higher BFE than the 75 used:25 virgin blend suggesting a non-linear relationship and highlighting that performance is influenced by factors other than virgin content.
Implications for print performance
The preceding data show that used, virgin and blended materials can be robustly differentiated in terms of flowability via dynamic testing and that the properties of blends cannot be easily predicted. Powders with a higher BFE are more resistant to flow under the low stress, forced conditions applied during the test, conditions that are comparable to the those applied by the recoater as it spreads powder across the build platform. BFE is a useful metric for AM applications precisely because of this relevance and the sensitivity of the technique.
However, BFE alone cannot necessarily capture the full impact of passage through a printer, no single powder property can. One of the most important lessons we have learned about testing powders is the importance of matching test and process conditions. In an AM printer, powders are subjected to a range of conditions which means that multiple aspects of behaviour are likely to be relevant. For example, in the feed hopper the powder is subject to the consolidating force of its own weight, prior to discharge; packing efficiency in the powder bed ranks alongside speed of dispersion across the build platform when it comes to high throughput, high quality printing; and air pockets inhibit heat transfer and can translate into porosity and mechanical weakness in a finished part. A full assessment of the impact of re-use therefore needs to include measurements such as bulk density, and response to air, how easily the powder releases air and settles into a densely packed bed.
One of the attractions of instrumentation for dynamic testing is that it offers complementary capability for bulk and shear property measurement. These capabilities allow multi-faceted characterisation of re-used powders and, ultimately broaden the scope when investigating the relationships between measurable powder properties and printing performance.
Results from a study carried out by researchers at the National Center for Additive Manufacturing Excellence, Auburn University using a commercial 17-4 PH SS powder (LPW Technology, UK) demonstrate the value of this approach3. Batch 1 is virgin powder with subsequent batches produced by sieving (80 µm screen) residual powder from the preceding print run.
Bulk density data indicate that particle packing efficiency increases steadily, though not linearly, with re-use. This suggests that re-used powders may actually be preferred in terms of finished component porosity. Aeration Ratio (AR) is a dynamic property, the ratio of BFE measured using a conditioned sample to that measured with air flowing through the sample at a controlled velocity. Aeration exerts a separating effect on powder particles and AR is therefore strongly influenced by inter-particular forces such as cohesion.
Mechanical testing reveals that AR values reflect a trend in ductility, as assessed via uniaxial tensile testing, where RA% is reduction in area and Ɛf is true strain at fracture. Additional testing4 highlights the superior fatigue strength of Batch 15 at high cycle frequency, a result attributed to lower porosity that can be correlated with bulk density.
Here then, dynamic and bulk properties both proved relevant to print quality. It is also worth noting that shear properties did not. Cohesion data failed to differentiate the more heavily re-used samples (see graph).
Anyone working in the AM industry will readily recognise the difficulty of robustly characterising used powders to make secure decisions about re-use. Indeed, these difficulties reflect those associated with specifying AM powders in the first instance, as previously discussed.
For those tackling this challenge there are some important take-aways from the work highlighted here. One is the sensitivity and relevance of dynamic properties which have repeatedly proven illuminating with respect to AM performance. The other is the value of complementary test protocols, notably bulk property measurements which help to elucidate packing efficiency, a defining characteristic of AM powders. Having more than one testing protocol to hand is highly advantageous when optimising AM powder use.
Acknowledgements and thanks:
The National Center for Additive Manufacturing Excellence, Auburn University and LPW Technology (Carpenter Additive).
1 J. Clayton and R. Deffley ‘Optimising metal powders for additive manufacturing’ Metal Powder Report, Sept/Oct 2014
2 ‘Powder Flow Testing for Additive Manufacturing’ eBook available for download at: https://www.freemantech.co.uk/learn/ebooks
3 P. D. Nezhadfar et al ‘The Effects of Powder Recycling on the Mechanical Properties of Additively Manufactured 17-4 PH Stainless Steel’ Solid Freeform Fabrication 2018: Proceedings of the 29th Annual International Solid Freeform Fabrication Symposium – An Additive Manufacturing Conference.
4 A. Soltani-Tehrani et al ‘Fatigue behavior of additively manufactured 17-4 PH stainless steel: The effects of part location and powder re-use.’ Additive Manufacturing, 36, 101398.
Read more from Freeman Technology’s Jamie Clayton:
Want to discuss? Join the conversation on the Additive Manufacturing Global Community Discord.
Get your FREE print subscription to TCT Magazine.