Selective laser melting: Impact of powder on steel components porosity

While work is carried out to describe powder characterisation techniques and to establish those most suitable for AM processes, little has been done to understand and review the relationship between powder properties and process signature or final part quality. Here we review the impact of particles characteristics on the porosity of steel components built using SLM.

Without clear guidelines regarding suitable requirements for additive manufacturing powder requirements, different providers setup with various equipment supplies various powder properties.

a/b: gas atomisation - d/e: rotary atomisation - g/h: Plasma rotating electrode process [1]

In addition, during processing parameters such as layer thicknesses vary (20to100um) depending on the machines used, the parts geometry, or the in-house parameters development procedures influenced by productivity rate. We report preliminary results assessing the impact of different powder size and size distribution on the density of steel components.

Powder layer density

For SLM, the powder layer density should be as high as possible [2] in order to produce dense parts with high scan velocities and high productivity. The density of a powder layer is dependent on the particle sizes or size distributions as well as the recoating blades.

Smaller powder sizes with higher relative powder densities require less energy to sinter. A wider distribution of particles sizes can also allow for higher powder density, since smaller particles can fill the gaps between larger particles. McGeary [3] demonstrated that specific ratios of bi-modally distributed powder sizes can achieve an optimal packing density of 84 % with a 1:7 size ratio and a 30 % weight fraction consisting of the smaller size. 

However, if the amount of fine particles, (<6/10μm) is too high, the agglomeration of particles can eliminate their positive effects of filling up voids. These particles can present health and safety risks as their handling trigger creation of 'dust' cloud. 

Component density

The role of powder size and size distribution affects the relative density of the powder, which in turn affects the activation energy required for heated particles to coalesce [4], [5]. The factors influencing final part density are  energy input required to melt powder completely (P, s, h) and the radiation time over a specific localised area which affects the melt lifespan, the maximum melt pool temperature and its cooling rate.

Theoretical layer thickness = 30um (effective layer thickness =47um)

For lower layer thickness, coarser powders typically result in higher porosity (at fixed energy density) due to:

• Energy input insufficient to fully melt the larger particles (not fully irradiated) 
• Partially unmelted coarser particles generates voids within the scanned layer 
• Emerging surface roughness promotes inability to fill the valleys between the scan tracks of the last layer

For equally coarse powder, the amount of fine particles explains higher density behaviour for larger layer thickness due to trickling down between larger particles and promoting sintering.

Theoretical layer thickness = 45um (effective layer thickness =74um)

For a fixed energy density or scan speed, density is about 1% lower for all three powder types than for 30um layer thickness due to:

• the energy flux is reduced quickly with increasing layer thickness
• when effective layer thickness is >> theoretical layer thickness, amount of fine powder grains becomes comparatively less important.

Take away

Preliminary study on 316L powders show that for high density:

• the effective powder layer thickness teff  must be at least 50% higher than the diameter of 90% of the powder particles so that most of the particles can be deposited within teff:  teff > 1.5 x D90
  
• Sufficient amount of fine particles are necessary to fill the voids between the coarser grains. D90 > 5x D10
• Both requirements together indicate that D10 is about 7.5 times smaller than teff

References:
[1] A. Strondl, O. Lyckfeldt, H. Brodin, and U. Ackelid, “Characterization and Control of Powder Properties for Additive Manufacturing,” Jom, 2015.
[2] Karapatis, N.P., A sub-process approach of selective laser sintering. 2002, Ecole Polytechnique fédérale de Lausanne EPFL: Lausanne.
[3] Karapatis, N.P., G. Egger, P.-E. Gygax, and G. Glardon. Optimization of Powder Layer Density in Selective Laser Sintering. in Proc. Of the 9th Solid Freeform Fabrication Symposium. 1999. Austin (USA).
[4] I. Robertson and G. Schaffer, “Some effects of particle size on the sintering of titanium and a master sintering curve model,” Metall. Mater. Trans. A, vol. 40, no. 8, pp. 1968–1979, 2009.
[5] H. Su and D. L. Johnson, “Master sintering curve: A practical approach to sintering,” J. Am. Ceram. Soc., vol. 79, no. 12, pp. 3211–3217, Dec. 1996.
[6] Comparison of density of stainless steel 316L parts produced with selective laser melting using different powder grades, A. B. Spierings, G. Levy, SFF Symposium 2009 Review Paper