Centrifugal Atomization Methods

In centrifugal atomization, centrifugal force breaks up the liquid and throws off the molten metal as a spray of droplets that then solidify as powder particles. As shown in Fig. 1, there are several different types, which are discussed. In general, centrifugal atomization methods are far more energy efficient than gas and water atomization, where only about 1% of the jet energy is used in the disintegration of the metal stream (Ref 1). In contrast, the energy used in centrifugal methods is low as all the rotational work directly accelerates the metal droplets, instead of the atomizing medium, as in the case of two-fluid techniques. Centrifugal atomization also generally leads to a much narrower spread in particle sizes than does gas atomization (see Fig. 2), with rrg as low as 1.2 to 1.4 in some cases.

As the process depends on the solution to the problem of finding a compatible material for the spinning disk or cup, the applications that have been, and are currently, used on a significant industrial scale are quite distinct. They are reviewed in order of ascending melting point of alloy.

Spinning Disk Atomization of Electronic Grade Solder. Solder powder for electronic applications has a very demanding specification; it must be perfectly spherical and satellite-free, it must be very low in oxygen content 100 ppm), and it must have a very narrow size distribution. In 1997, more than half of the demand was for type III size grade that is nominally -45+25 /'m. and often demanded as -40+30 ¿'m. Attempts to make this product with inert gas atomization have now virtually ceased as yields are as little as 5% and the avoidance of satellites is very difficult.

In the United States, Japan, and Europe, there are many producers using spinning disk methods to make this product. This is possible because a steel disk is well wetted by normal Sn63Pb37 solder and is not eroded very fast. A disk with diameter of 40 to 100 mm running at speeds of 30,000 to 60,000 rpm can produce this material with good yields of 30 to 70% and at (gross) rates of 50 to 100 kg/h. The plant is only about 2 to 3 m in diameter and inert gas filled. Unfortunately, a small amount of ultrafine particles is produced as a secondary peak in the distribution, and some problems with satellites are found. Annual production, if the modest yields are taken into account, is probably currently several thousand tons per year.

Spinning-Cup Atomization of Zinc, Aluminum, and Magnesium. When used on higher-melting metals, it is difficult to run at the sort of speeds that are used on solder. However, there are markets for coarser powders of zinc (alkaline batteries), aluminum (chemical), and magnesium (flares) that have made this a significant process. In all cases, the cup is 100 to 200 mm diameter, running at moderately high speeds, from 3,000 to 10,000 rpm. The vessel size needed is very large, up to 12 m in diameter, but productivity can be very high. In the case of zinc for battery applications, which is commonly required as -600+100 t'm material, air atomization may give yields of 70 to 80%. Using a 5 kW spinning cup, 98% can be achieved, at outputs of several tons per hour, and with no compressor costs (an equivalent air atomizer might use several hundred kilowatts of compressed air). However, large output is needed to justify the large-scaled plant needed. Many thousands of tons per year are made in this way.

Aluminum is also processed in this way, but because of the large size of plant needed to freeze coarse particles, it is done in open air, which produces needles (see Fig. 3d). In this case, a perforated steel or cast iron cup is used to make a series of streams of metal, which break up into needles due to the oxide film on them. Production is considerable, probably thousands of tons per year.

Magnesium powder for flares was made in several similar systems in the United States during the Vietnam war at rates of 1 ton/h. Excellent yields of a powder almost free of the very dangerous -100 /'m fines were obtained. In this case, the fact that liquid magnesium does not attack iron simplified cup manufacture. Due to lack of current demand, this production process is seldom used.

The rapid solidification rate (RSR) process, which was first developed in the 1970s by Pratt and Whitney for making superalloy powders, is another form of centrifugal atomization. To overcome the problems of the material in handling high melting and aggressive alloys, the process employs a high-speed water-cooled rotating disk (20,000 to 30,000 rpm), which breaks up the molten metal stream. To enhance solidification rates, the resulting droplets are then hit by high-pressure helium gas as they leave the periphery of the rotating disk. The RSR powders are spherical, with an average particle diameter <100 /,!m. High cooling rates (>105 °C/s) are achieved in the small powder particles, and this leads to a high degree of compositional homogeneity and fine-scale microstructure. An early plant of this type is shown in Fig. 28.

Centrifugal Atomization Process

Fig. 28 Schematic of first-generation RSR machine. The second-generation machine incorporates closed-loop helium recirculation, higher atomization speeds, and a three-fold increase in melt capacity. The third-generation machine retains these features, but the melt capacity is increased to 900 kg (2000 lb).

Fig. 28 Schematic of first-generation RSR machine. The second-generation machine incorporates closed-loop helium recirculation, higher atomization speeds, and a three-fold increase in melt capacity. The third-generation machine retains these features, but the melt capacity is increased to 900 kg (2000 lb).

The first RSR unit was operated in late 1975, producing IN-100 powder as a model, and patents were issued in 1978 (Ref 37) and 1982 (Ref 38). In addition to superalloys, the process has been used to produce specialty aluminum alloy, beryllium alloy, molybdenum, titanium alloy, and silicide powders. The RSR powders currently produced by Pratt and Whitney are nickel-base superalloys, steels, and aluminum alloys. These alloy types are induction melted in vacuum. In a more recent development of the RSR process, reactive metals (titanium, molybdenum) and silicides have been arc melted prior to atomization. For titanium alloys, melt sizes up to 45 kg have been atomized successfully.

For nickel-base alloys, representative operating conditions are disk rotation speed 2500 radians/s, melt flow rate 0.2 kg/s, gas (helium) velocity 170 m/s, gas mass flow rate 1 kg/s, helium backfill partial pressure 33 kPa, and melt superheat 100 °C (180 °F). A major consideration in the design of this rotating disk atomizer was the desire to achieve extremely high cooling rates in the droplets by convective cooling in the helium gas. Depending on the diameter of the rotating disk, droplets are ejected at velocities in the range 40 to 110 m/s. Typically RSR droplets smaller than 100 /'m diameter cool at rates of about 105 °C/s.

This technique produces notably narrow distributions and is used commercially to make plasma spraying powders where size ranges such as -150+53 or -100+38 /'m may be needed. However, the uncontrolled geometry of the skull on the rotating disk, and the problems of out-of-balance forces, degrades the size distribution significantly compared with that achieved with lower-melting materials where the cup or disk has precise geometry. It also results in heavy maintenance costs on the spinning-cup assembly. As the melting process is normally batched, this is a major drawback. Lower-melting materials are easily supplied continuously, avoiding or minimizing startup and maintenance problems. As a result of these disadvantages, gas atomization has continued to dominate the production of higher-melting alloy powders. The use of huge volumes of costly helium is another drawback, and the benefits of the rapidly solidified microstructures have yet to find major commercial paybacks outside the aerospace sector. Thus, Pratt & Whitney remains the only operator in the 1990s, and the few attempts that have been made to develop similar devices elsewhere have not been productive. Production is probably less than 1000 tons/year.

+1 0

Post a comment