381719 Powder Characterisation for Powder Metallurgy and Additive Manufacture

Friday, November 21, 2014: 8:30 AM
210 (Hilton Atlanta)
Tim Freeman, Freeman Technology Inc., Tewkesbury, United Kingdom, Mike Delancy, Freeman Technology Inc., Medford, NJ, John Yin, Freeman Technology Inc, Bayside, NY and Katrina Brockbank, Freeman Technology Ltd, Tewkesbury, United Kingdom

Manufacturing of products from a powdered substrate is well established within the metallurgical industries.  Processing usually involve the dense packing of a powder into a die or mould which can then be subjected to sintering through compaction and/or heating. 

More recently the technique of Additive Manufacture (AM) has also made use of powder materials to create products through layer by layer deposition and precise, localised fusing of a feedstock to itself through, for example, laser sintering or electron beam melting.

Although standard sintering techniques are sensitive to the flow behaviour of their feedstocks, AM is arguably even more sensitive to variability in powder performance due to the need to apply fine layers of powder in a uniform manner.  This therefore demands an even higher degree of consistency and repeatability in the powdered feedstock(s).  Even the slightest perturbation in the evenness of each layer caused by, for example, intermittent flow from the hopper/shoe or the presence of voids and agglomerates will create a surface disturbance/discrepancy that can result in a localised component weakness, or surface blemish that will render the component unusable.

Many users rely heavily on a consistent particle size distribution of their powders as a critical quality attribute (CQA), but in many cases the size distribution alone , or indeed any other single parameter (such as Hall Flowmeter or Hausner Ratio), is not sufficient to fully qualify a batch of feedstock.

The complexity of most powder manufacturing demands a thorough understanding of the nature and behaviour of the powders themselves.  It is entirely inadequate to suggest that a powder’s characteristics can be represented by a single number from a single test [1].  Indeed the manufacture of different AM products invariably uses a range of equipment configurations for powder processing storage and transport/transfer – each one may be sensitive to one or more properties of the bulk powder feedstock.  Therefore knowing how a powdered material will behave over a range of stress conditions – when stationary, in motion or about to move – is vital for designing and running AM processes.

In this paper, we review the properties of a number of powdered metals used in AM applications and establish how certain testing responses can be related to variations in their processability.  The study is presented in two parts.

Firstly, three batches of a Stainless Steel powder known to behave differently in AM, but difficult to differentiate using conventional methods, were tested using a multivariate approach [2].  The three batches (A, B and C), which had identical particle size distributions and Hall Flowmeter results, exhibited different behaviour when dispensing from a hopper: batches A and C showed acceptable behaviour across the process, while batch B showed very poor behaviour, routinely blocking the dosing mechanism and displaying variable properties in the final product.  Clear, repeatable differences were identified between the samples, directly correlating with their process performance.  Importantly, the key differences in behaviour between the samples – batch B had significantly greater compressibility and much lower permeability than the other two samples – also provided a rational explanation of its poor processability.  Thus it was possible to establish a design window of key performance characteristics for acceptable process behaviour, and quantified differences between the powders that other techniques cannot reliably identify. 

Secondly, the properties and behaviour of three samples of a metal alloy powder (with a Gaussian size distribution between 40 and 200microns) were evaluated – two manufactured using gas atomisation but from different suppliers and a third generated using an alternative process.  Whilst the size distribution of the resultant powders from both processes can be closely controlled, the range of other particle properties (particle shape and shape distribution; surface texture; surface chemistry effects – hydrophobicity/hydrophilicity and tribo-electric potential, etc.) will all contribute to the flow behaviour of the bulk powder.  Whatever manufacturing method is used, it is entirely possible that what is ostensibly the same process may, in fact, produce powders which flow in considerably different ways.  This can be due to subtle or gross variations in the entire production train – aspects such as heating and cooling rates; in-plant handling and transport variations; storage times and stress levels; may all affect powdered product.  The full evaluation of these samples will therefore allow the comparison of supplier variation as well as establishing the ramifications of processing materials generated by different manufacturing techniques. 

Significant differences were observed in all of the flow properties.  Considering the two gas atomised samples, one showed significantly higher resistance to dynamic flow and much lower permeability, whilst their shear results were identical.  The plasma atomised sample was much freer flowing than either gas atomised samples in both dynamic and shear properties. 

In both studies, it can be seen that samples which are considered physically identical (due to the same particle size distribution and/or Hall Flowmeter results) demonstrate significantly different flow behaviours.  In the first example, these disparities translate to variation in process performance which in turn leads to sub-optimal product quality and/or throughput.  Such variability can be difficult to accommodate during processing and can lead to process interruption and/or reduced product quality.  In the second case, it is clear that manufacturing differences – whether from different suppliers of from alternative manufacturing techniques – can significantly alter the behaviour of a material even though the particle size distribution remains the same.  These results demonstrated that not only can the differences in the manufacturing be quantified, but also that the impact of changing suppliers can be understood.

The important implication is that using particle size distribution alone as a critical quality attribute to assess the suitability of a powder can result in significant processing issues and reduced output quality. 


[1]   Freeman R. 2007. Measuring the flow properties of consolidated, conditioned and aerated powders – A comparative study using a powder rheometer and a rotational shear cell. Powder Technology 2174 (1-2):5-33.

[2]   Prescott, J. K. & Barnum, R. A. 1984. On Powder Flowability. Pharmaceutical Technology, 60-84.

Extended Abstract: File Uploaded
See more of this Session: Solids Handling and Processing II
See more of this Group/Topical: Particle Technology Forum