Powder Comparison: Standards for DED Material Substitutes

Editor’s note: This piece is based on research conducted by Jamie Clayton of Freeman Technology, Lisa Markusson, previously of GKN Aerospace Sweden AB and Lars Östergren currently of GKN Aerospace Sweden AB. 

Directed energy deposition (DED), also called laser metal deposition (LMD), uses lasers to fuse materials as they are being deposited. This mid-process fusion differentiates DED from binder jetting and powder bed fusion while also making the technique suitable for repair work. DED causes lower thermal stress than arc welding or thermal spraying, and the ease of switching between materials when using this technique improves its versatility.

Optimizing DED relies on selecting powders that will process well and produce components and repairs with desirable properties. Many of these powders are developed specifically for DED — so what can users do when supply chain demands leave them with shortages of powder?

Researchers at Luleå University of Technology and GKN Aerospace Engine Systems set out to discover which particle and powder properties correlate to similar DED performance and build qualities. They compared five Inconel 718 powder samples — one powder which GKN regularly uses (powder C), as well as four others (powders A, B, D and E) — from four different suppliers.

The Experiment

To assess the print performance and deposit quality of each powder, the researchers printed single and multi-bead deposits onto Inconel 718 plates. Process parameters were optimized for powder C and left the same for each powder. Measurements of the powder supplied versus the weight of the deposit determined powder efficiency; while defects, geometry, and microstructure determined deposit quality.

While all four alternative powder supplies offered higher efficiency than powder C, the higher variability of powder E made it far less desirable. In respect to bead quality, pore number and pore size, powder A produced results closest to powder C. From here, the researchers compared several physical characteristics to determine which properties could differentiate powders A and C from the others.

Before beginning research, GKN’s specifications for DED powders referenced composition, manufacturing method, Hall Flow Index and particle size distribution. Over the course of the study, however, particle size and manufacturing method proved unable to successfully predict powder A as the closest performance match to powder C, and while Hall Flow Rate suggested the two were similar, all five powders still passed GKN’s established specification on the parameter.

Porosity in the feedstock powder appeared to be better linked to porosity in the deposit, and powders A and C were the closest analogues in specific energy (which quantifies unconfined flow behavior). Similarities in the difference between the powders’ conditioned bulk density and bulk density after 50 taps of vibrational consolidation also distinguished powders A and C from their counterparts.

Conclusions

Ultimately, the researchers concluded that powder efficiency was correlated with basic flowability energy and deposit pore frequency was correlated with particle pore frequency. While they acknowledge the limited data set, they also argue that a small selection of highly similar powders is representative of the practical challenge DED users face when assessing alternative supplies.

Therefore, they say most users can learn from four primary takeaways. First, dynamic testing provides far greater insight into flowability behavior and provides more robust differentiation between samples than the Hall Flow Index. Second, flowability directly impacts performance metrics such as powder efficiency. Third, part quality — especially bead porosity — is highly dependent on the density and porosity of powder particles. Lastly, particle shape influences powder suitability.

For a new metal powder specification, the researchers say any new metal powder specifications should include notes on basic flowability energy, along with a descriptor of particle morphology and particle density.