Can powder testing accelerate the development of new feedstocks for AM?

Current commercial applications of powder-based AM predominantly use well-established materials such as polyamides, titanium, or stainless steel, in the form of feedstocks tailored for printing. However, flexibility is a defining feature of AM and there is almost limitless potential to print with new materials. Across the globe researchers are working with nanocomposites, filled and reinforced polymers, innovative alloys, and other materials, to add new functionality to AM parts and realise new applications for the technology.

These researchers share a common problem which is how best to characterise new materials to assess their likely performance. Developmental powders may only be available in very small quantities, ruling out a print trial, but gaining a representative assessment of powder behaviour, as early as possible, is critical. This raises the question of whether it is possible to test powders, in the absence of or ahead of a print trial, to accelerate development to a successful conclusion?

My view would be that yes, an optimised testing strategy can make a big difference when it comes to developing new powders for AM. While there is no substitute for a print trial, sensitive, relevant testing can help to narrow down the candidate pipeline, to point research down the most promising avenues, and to provide insight into why one powder ultimately turns out to be superior to another.

Focusing on the physical, beginning with particle morphology…

Example data from a study carried out by researchers at the Centre for Additive Layer Manufacturing (CALM) based at the University of Exeter (UK) [1] exemplifies how physical testing can help accelerate innovation in powder development. In this work, novel nanocomposite powders for laser sintering were synthesised by encapsulating graphene nanoparticles (GNP) on the surface of polyamide 12 (PA12) particles in a thin layer of poly (vinyl alcohol) (PVA). The aim was to demonstrate the feasibility of using a fast and cost-effective method of fabrication for nanocomposite powders which eliminates any health and safety risks related with handling large quantities of polymer and loose nanomaterials. The printed parts would impart thermal and electrical conductivity benefits in addition to lightweight and mechanical performance. 

Measurements of particle morphology are typically the starting point when it comes to the physical characterisation of powders. The plot above shows particle size distribution (PSD) data for four samples – the original PA12 powder, PA12 coated with PVA, and two samples with GNP encapsulated into the PVA coating, at 0.1 and 1 wt% respectively. These results indicate that the coating/encapsulation process has little impact on particle size.

Further morphological information was gathered by applying dynamic image analysis, scanning electron microscopy (SEM), and transmission electron microscopy (TEM). These data quantified the circularity and sphericity of the powder particles. Similarly to PSD data, the results indicated that coating minimally impacts particle morphology with the thickness of the coating observed to be only around 1µm; GNP particles were successfully and uniformly embedded. The only real difference these analyses detected was in surface roughness with coated particles appearing smoother than the original PA12 material.

When it comes to the physical assessment of AM powders there are two distinct areas of interest:

  • How the powder flows and spreads – since this influences the ease and speed of layer deposition, a vital step in both binder jetting and powder bed fusion processes.
  • How the powder packs – since this impacts the uniformity and consistency of powder layers and improves heat transfer within the powder bed.

Particle size data provide some insight into these behaviours since particle size and distribution directly influence packing behaviour and flow properties. Particle shape data is also relevant. For example, powders consisting of smoother, more spherical particles typically flow more easily than those that are rougher and/or more irregular.

The measurements of particle morphology presented here therefore hold valuable information and the results are encouraging with respect to producing a material that prints at least as well as the base material, an established AM feed. However, in isolation these results do not provide all the information required. Is the difference in surface roughness significant in terms of how the powder will flow or pack? And have other, potentially important particle properties, such as surface charge, changed in a way that these techniques fail to detect?

When assessing how powders will flow and pack particle properties alone typically cannot provide the answers. Adding bulk powder property measurement to the tool set is the way forward.

adding in bulk powder properties.

The figure and table below show a selection of dynamic flow properties for the four samples, along with conditioned bulk density (CBD) data, all measured using an FT4 Powder Rheometer (Freeman Technology, Tewkesbury, UK). Basic Flowability Energy (BFE) and Specific Energy (SE) both quantify the ease with which a powder flows, under confined (forcing) and unconfined (low stress or gravitational) conditions, respectively. Significant differences between the powders are observed. Coating improves flowability, reducing both BFE and SE. Encapsulating GNP within the coating further extends this effect, though there is minimal difference between the results for 0.1 and 1% wt% GNP. The smoother surface of the coated particles, and a corresponding reduction in particle friction, provides a rationale for these results.

CBD is a useful metric for assessing packing performance with higher values indicative of closer, more efficient packing. Here, the samples are effectively indistinguishable, despite the precision of the measurements, suggesting that any changes in morphology have little impact on this important characteristic.

Turning finally to Stability Index (SI) which is determined by comparing initial BFE values with those measured after a seven repeat test cycles (as shown in the plot), we can see that all the samples exhibit good physical stability (SI close to 1). Changes in BFE with repeat testing are a sensitive detector of issues such as particle break-down/attrition so it is particularly helpful to see the exemplary stability of the two GNP composite samples.

Viewing this data through the lens of answering the questions needed for powder development its clear that we have added considerably to what we knew from particle properties. The composite powders have better flow properties than the base PA 12, under both forcing and low stress conditions suggesting they will be easier to print with. This can be viewed as a relevant quantification of the impact of the observed changes in morphology, notably surface roughness. Furthermore, the powders are physically stable, there is no observable break-down of the coating with repeated shear. Packing efficiency is similar to the core feed.  

These data suggest that the composites will behave at least as well as the PA12 and essentially provide a green light for further investigations, for print trials and on to evaluation of the properties of printed components.

Validating relevance

In this study the 0.1 wt% GNP nanocomposite powder was selected for print trials in a commercial laser sintering printer (targeting superior mechanical properties rather than thermal/electrical conductivity) and went on to perform well, as would be expected from the physical characterisation data. Furthermore, the mechanical properties of the printed test specimens were found to be superior to those of analogous PA12 samples.  

For these researchers, the translation of encouraging flowability values into good process performance comes as no surprise since this set of physical tests is an established approach. The forcing conditions applied during BFE measurement mean that it can be strongly influenced by CBD (not a complicating factor here) so trends require careful evaluation, however, over multiple studies this team has consistently found BFE values to be relevant to print performance [1 – 3]. As a result, BFE, along with other dynamic properties, is trusted as a reliable primary screen for increasingly sophisticated powders.

Such powders now include metal-organic frameworks (MOFs)/PA 12 nanocomposites produced by an in situ hydrothermal synthesis which grows the MOF (in this case zeolitic imidazolate frameworks – ZIF 67) directly on to the surface of PA12 particle [3]. MOFs are a relatively new class of crystalline solid materials with extremely high specific surface area.  This makes them attractive for applications such as carbon capture and gas storage.  ZIF-67/PA12 nanocomposite powders exhibit lower BFE and SE values than the core PA12 powders, as with the GNP nanocomposites, confirming their processability. The CALM team has used them to print lattice structures with controlled cavities and macro-pores that optimise access to the high surface area of the MOF, thereby enhancing CO2 adsorption performance.

This exciting project highlights the potential to combine AM with cutting edge material science to develop solutions to demanding applications. The right physical testing techniques have an important role to play in accelerating such development.


[1] B. Chen et al. ‘Laser sintering of graphene nanoplatelets encapsulated polyamide powders’ Additive Manufacturing 35 (2020) 101363

[2] B. Yazdani et al ‘A new method to prepare composite powders customised for high temperature laser sintering’ Composites Science and Technology 167 (2018) 243- 250

[3] B. Chen et al. ‘In-situ synthesis of Metal Organic Frameworks (MOFs)-PA12 powders and their laser sintering into hierarchical porous lattice structures’ Additive Manufacturing 38 (2021) 101774

Want to discuss? Join the conversation on the Additive Manufacturing Global Community Discord.  

Get your FREE print subscription to TCT Magazine.

Source link

Show More

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *