Particle Size Distribution

When various mechanisms operate to produce a powder, these occurrences should appear as peaks in the frequency distribution plots. Nitrogen and argon atomized powders in Fig. 11 show three areas of peaks. The first, at 5 /'m. is considered to represent the daughter particles (Ref 21, 22). The third peak belongs to the stable sizes, which range from 30 /m to 100 /''m and greater. The intermediate peaks are taken to indicate considerable overlapping between the daughter particles of large droplets breaking up at lower relative velocity and stable sizes of initially small drops shattering at high relative velocity. The peaks are not so well defined in the case of helium atomized powder, showing complete merging.

Fig. 11 Frequency distribution for AA 2014 powders made using helium, argon, and nitrogen. Source: Ref 20

The distribution of particle sizes in atomized metal powders is often found to comply approximately with the log-normal law (Ref 21, 22). The two mechanisms for the production of fine and coarse droplets described above are not compatible with this law, which is strictly unimodal. The wide range of sizes covered is its main advantage in representing satisfactorily the size distribution of gas-atomized metal powders.

Several mean diameters are used to describe the size of a powder. The most common is the mass median diameter, dm, i.e., the 50% point on the cumulative weight versus particle size graph. Two other useful mean diameters are the Sauter mean diameter dvs and the volume mean diameter dvm, defined by:

6/vs = Ex\/A7Ex26/A' = 100/(Ed$/x) tlym = SAW/SAW = ^xd$/100

where 1x = mean diameter in a size band, dN = number of particles in size band, d<P = weight percent in size band. Sauter mean diameter is the diameter of a sphere with the same surface area per unit volume as the powder and is sensitive to changes in the fine particle range of the powder. Volume mean diameter, on the other hand, is a moment mean, sensitive to changes in the coarse particle range. Both these two diameters are easily calculated from the sizing data and are routinely reported in sizing reports obtained in light scattering type instruments.

A single mean diameter is not sufficient to describe a powder; information is also required on the spread of sizes about the mean. This is measured by the geometric standard deviation of the log normal distribution and is calculated:

The irg values for gas-atomized aluminum powders tend to increase as the mean particle size increases indicating that the distribution becomes less tight for coarse powders. The typical range of ng is from 1.8 to 2.5 (Fig. 12).

S 22

S 22




D jgB




Mass median diameter, nm

30 40

Mass median diameter, nm

Fig. 12 Variation of geometric standard deviation with median particle size for as-atomized particles

In the case of log-normal distribution of sizes, all mean diameters are related to each other. Hence, if one diameter and the value of g are known, all other mean diameters can be calculated. The relationship between the two common mean diameters, dvs and dvm, and the mass median diameter, dm, is given by the following equations:

ln dvm = ln dm + 1/2ln In dvs = ln dm - l/21n2 dL-dvsd


These relationships are useful in checking consistency of the data and the applicability of the log-normal law to any size distribution obtained. Figure 12 can be used for quick estimates of the g from the ratio of these mean diameters.

Control of Particle Size. By modifying process parameters and screen sizes, the particle size distribution of atomized powders can be adjusted to produce a wide range of products from aluminum granules (typically +200 mesh) to fine atomized powders (up to 99%, -325 mesh). Figure 13 shows particle size distribution curves for typical coarse, medium, and fine grades of atomized powders. Besides the wide range of grades available from direct atomizing, others are possible from blending screen fractions and air classification.

Particle diameter, ^im

Particle diameter, ^im

Fig. 13 Particle size distribution measured by Microtrac in some Alcoa powder grades (grade 130/2 is by sonic sieve method)

Important process parameters in atomizing pure metals include nozzle configuration, nature, pressure, and temperature of the atomizing gas and delivery rate, and temperature of the molten metal. For alloy powders, nature and content of the alloying elements also need to be taken into account. Special precautions are often needed in atomizing alloys that contain alloying elements that evaporate or oxidize readily.

The geometry of a nozzle controls the flow of the atomizing gas and is therefore of the utmost importance in any atomization process. Information on the influence of nozzle design is scanty and empirical, and it mostly originates from patent literature. A recent study on a confined design nozzle has amply illustrated the influence of nozzle configuration (Ref 5) on the particle size of the as-atomized aluminum powders. Some results from that study are shown in Fig. 14. To eliminate variation due to differences in the amount of gas delivered by different designs, the results were replotted against the ratio of gas used per unit metal atomized. It is noted that a good design can produce substantially finer powders for the same amount of gas used. In subsequent gas flow studies, good nozzle designs have been found to maintain the supersonic nature of the jet farther in the jet plume than poorer designs, thus allowing the secondary breakup process to take place in a high-velocity gas jet.






_• A



-O .


Fig. 14 Variation of mass median diameter for aluminum powders with the ratio of metal/gas

The influence of nature and pressure of the atomizing gas on particle size of the as-atomized powder are well established. Helium produces much finer powders than those produced by nitrogen and argon under similar conditions using the same nozzle (Fig. 15). Comparison on the basis of volume of gas used per unit weight of powder produced also shows helium to be superior to argon. Helium atomized powders as fine as 13.5 m median diameter have been obtained using unheated gas. Under similar operating conditions, nitrogen atomized powders were 23 m and argon atomized powders were 25 m. This difference is attributed to the higher gas velocities attained in helium in supersonic flow. Similar benefits are expected from higher gas temperatures due to increased velocities. Further, heated gases will help reduce the danger of the premature freezing of the metal at the nozzle tip. With respect to gas pressure, it is known that a given nozzle, when all other parameters are constant, will produce finer powders as the gas pressure is increased. However, as the gas flow rate also increases linearly with pressure, the observed effect is similar to a more favorable gas/metal flow ratio, as well as that of increased pressure. When the results are compared on a gas/metal flow ratio basis, higher operating pressures are still beneficial.

Fig. 15 Comparison of argon and helium gases on the basis of gas volume used per unit weight of powder produced

During manufacture, particle size control in the as-atomized state is achieved by the selection of the rate of atomization (metal delivery rate). For any given nozzle configuration and operating parameters, the rates have to be slowed down in order to make finer powders. The median diameter is found to be approximately proportional to the square root of the rate of atomization. This may be more generally expressed as a function of the metal/gas flow rate ratio (Fig. 14). Superheat in the molten metal was found to have only a small effect on the powder size providing that there was sufficient heat in the liquid to prevent premature solidification. This is reasonable, as any change would come about primarily through changes in such properties of liquid metal as surface tension, density, and viscosity. These properties are only mildly dependent on temperature.

As metal flowrate is increased and the powder becomes coarser for any given atomizing conditions, in general, the spread in powder sizes (as measured by g) also increases (Fig. 15). This spread has been taken as indirect evidence for the presence of stable particle sizes in the powder after the completion of secondary breakup (Ref 5, 7, 8, 21).

+1 0

Post a comment