Particle Size Distribution

The factors affecting crg are somewhat debatable and poorly understood. The data shown in Fig. 11 (Ref 14) might lead to suggestions that very coarse distributions, toward the millimeter range, are tighter (lower <Jg). It is easily argued that extremely low pressures must give tighter distributions as liquid drops cannot be stable in the centimeter range. However, in the author's experience, there is little firm evidence linking ag to pressure where particle sizes are in the range from 10 to 200 m. While it is quite easy to design atomizers with too high an apex angle allowing rejection of the coarse droplets (which can give poor <Jg values up to well over 3.0), it is extremely difficult, indeed practically impossible, to make a jet system that produces a uniformly low value of <TS below 2.0 for a range of metals.

5 10 20 50 100 200 500 1000 2GD0 5000

Particle diameter, jim

Alloy and water pressure


Copper atomized at 54 MPa



18Cr-10Ni-2.5Mo stainless steel atomized at 50 MPa



Fe-15%Si atomized at 20 MPa



M2 high-speed steel atomized at 14 MPa



Fe-45%Si for welding electrodes atomized at 4 MPa



Zinc for alkaline manganese batteries atomized at 5.5 MPa



Copper shot atomized at 2.0 MPa



15% phosphor-copper shot atomized at 0.15 MPa



15% phosphor-copper shot atomized at 0.05 MPa


Fig. 11 Typical water-atomized particle size distributions. Source: Ref 14

As a general characteristic of water atomization, dm and rrg are not particularly sensitive to metal flow rate and the ratio of water flow to metal flow rate. Figure 12 shows that atomization of Fe-15wt%Si powder at rates from 20 to 110 kg/min (stream diameters from 5 to 13 mm) gives very uniform results.

20 30 60 SO 100 160 200

Particle diameter, ¿im

Fig. 12 Effect of metal flow rate on size of water-atomized Fe-15wt%Si powder. Source: Ref 15

The main influence on <7„. once a good and stable atomizing jet setup is ensured, appears to be melt chemistry. There is no satisfactory explanation of this yet published, but there is no doubt that alloys such as Ni-Cr-B-Si and Fe-Si(15-45 wt%) give consistently narrower distributions (&g "-4.6-1.8) than other metals such as copper, nickel, and iron. It can be speculated that this is related to the tendency of such alloys to spherical shape, in turn related to the characteristics of the oxide films formed on the particle surfaces. Those on "self-fluxing" alloys seem to be liquid below the melting point of the alloy and allow it to spherodize before freezing. As a particle of a given mass can appear on sieving to have various sizes depending on its shape, it may be that a spherical shape, which allows minimum variation in shape and thus reported size, favors low ffs. However this rationale is largely speculation that requires verification.

Figures 13 and 14 (Ref 16) illustrate the effects of nozzle diameter and pouring temperature for water-atomized copper powders on (Jg and dm. The large effect of the tank atmosphere—air versus nitrogen (Fig. 13) is interesting. In fact, it is known that if copper is atomized with a large dissolved oxygen content, the powder is much finer (about 30% reduction in c/m at constant pressure). Thus, the large difference between nitrogen and air atmosphere sizes may be due to this effect. Similarly, the chemistry may be directly influencing 1Tg. The generally lower values of Cg at lower metal flow rates reflect the importance of collisions in broadening the distributions.

Fig. 13 Effect of metal-stream diameter on geometric distribution of water atomized copper powder. Pouring temperature 1200 °C (2190 °F); figures in parentheses show average metal flow rates in g/s. Source: Ref 15

Fig. 14 Effect of pouring temperature on distribution of water-atomized copper powder. Metal stream diameter 4 mm (0.16 in.); water flow 320 L/min (85 gal/min); water pressure 13.2 MPa (1915 psi); atmosphere nitrogen. Source: Ref 11

The use of values of 0%, in combination with median particle size according to Fig. 11, permits the estimation of powder yields as a function of atomizing pressure. Where particles are needed in a narrow size range, for example, -150+53 /Jm, the impact of <7g on yields is very important (see Fig. 15).

100 90 80 70 60 S 50

30 20 10 1

Maximum/minimum size ratio

Fig. 15 Master yield curve showing yield between two sizes as a function of size ratio and standard deviation

0 0

Post a comment