108 Prakt. Met. Sonderband 47 (2015) 
processing methods on the microstructures of as-cast flakes and subsequent magnetic properties of VE 
the prepared magnets will be discussed. - z . 
2. EXPERIMENTAL PROCEDURE 
In order to satisfy the demand for high purity of Nd-Fe-B alloy, which is essential for obtaining 
good magnetic properties, the centrifugal atomization was performed in a inductively heated 
vacuum furnace. Special care was taken to avoid the contamination of the melt with oxygen and 
carbon. High purity base alloys (Nd-Pr, FeB, FeDy, GaDy) and metals (Fe, Al, Cu, Co) were used, 
with a purity of no less then 99,9 wt%. The nominal composition, which is presented in table 1, A 
was the same for all samples. SET 
Table 1. Nominal chemical composition with tolerances yom 
feB 
Chemical. Dy. Fe ..Nd mr. 8B Gi... A Cu G, SE 
compositio "36 64 wma 669302 6150153 er 
n +0,2 +0,5 +0,5 £0,1 £005 +0,05 +005 +0,05 +0,2 ig 
wioey 
In order to avoid oxidation, the vacuum furnace was vacuumed lower 10 mbar and then refilled micros 
with high purity Ar (5.0) three times. As the initial stage of purging the furnace was reached, {0 
oxygen analyzer, which is mounted to the Ar exhaust pipe showed less then 0,1 wt% of oxygen. ver 
The melting and casting itself was performed under protective Ar atmosphere. Special caution was bs 
given to the chemical stability of the crucible due to the aggressive behavior of the rare earth - 
elements in the molten state. A narrow chemical composition range is required in order to obtain (= 
good magnetic properties. Therefore no contamination of the melt in contact with the crucible can IF 
be expended. AO3 was chosen as a ceramic coating for the crucible and the tundish for its good he... 
chemical stability and proper adhesiveness to the crucible. An orifice of 10 mm was drilled in the 
bottom of the tundish and carefully painted with a thin layer of Al2Oz3. 
Both the crucible and the tundish were pre-heated. The crucible was inductively heated to approx. 
300 °C prior to loading of the raw material. The tundish was resistance heated with a heater which om 
was planted vertically into the tundish. Just moments prior to the casting stage, the vertical heater BET 
was lifted, thus preventing the instant solidification of the melt when it comes in contact with the 
tundish. The temperature of the tundish is measured by a thermocouple installed inside the heater 
and gives an approximation of the actual temperature of the tundish. The actual temperature is ARE 
approximated to 300 °C in the body of the tundish and 100 °C at the bottom, where the orifice is 
located. The cooling wheel frequency rate is regulated by a potentiometer. Frequencies of 210 Asa 
RPM, 240 RPM and 280 RPM were selected for the experiment. The orifice to cooling wheel & 
distance was set experimentally to 10-15 mm. Time of the melting stage was also varied in order to m 
achieve optimum homogeneity of the centrifugally atomized flakes. By extending the time of the i 
melting stage we assured that all of the raw material, especially Fe, with a high liquids temperature bes 
has melted. comp 
Table 2. Various melting conditions %) 
Sample 1  Sample2  Sample3 i 
RPM 7 210 240 280 5 
Melting time 6 10 10