were observed. These particles, however, were approximately twice as large as the particle size of the ingoing chromium powder. 3.3 Mechanical Working and Heat Treatment In order to promote homogenization the billet was hot extruded. To do this the billet was machined dry into a right cylinder and fitted into a low-carbon steel can. Machining was done under especially clean conditions. The extrusion can was sealed by electron beam welding after pumping down overnight. The canned billet was then heated in dry hydrogen to 1075°C in 2.5 hours, and extruded from 2-1/8" diameter down to 5/8" diameter (11.6/1 area reduction). The extruded product was annealed for 8 hours at 1300°C in dry hydrogen. This, however, failed to homogenize the sample as shown by metallographic examination. Similarly, further annealing at 1350°C for 15 hours failed to bring about the desired structure. The annealed extrusion was swaged to 0.275" at 850-950°c. A small piece was further swaged at room temperature from 0.275" down to 0.249". Both samples were annealed in pure dry hydrogen for (1) 72 hours at 1350°C and (2) 168 hours at 1400°C. 4. DISCUSSION Metallographic examination at the various stages of the processing has revealed that working and annealing failed to break up the chromium aggregates. The reason for the formation of these hard particles was not clear, especially since they were larger than the powders used in processing. It was noted that these materials appear to be closely associated with an oxide constituent also observed in the matrix. This is particularly evident in the 750X micrograph of the material in the extruded and swaged condition where it can be seen that, even though these particles had been severely deformed, an enveloping oxide film was still present (figure 1). The contamination is believed to be inherent in the commercial starting materials, since the test strips of clean stainless steel fired in the same furnace remained bright. It is suggested that although the overall furnace atmosphere was reducing to chromium, the oxide intimately associated with the chromium was not reduced because the powder billet densified readily leaving such areas cut off from dry hydrogen. Samples of this stainless steel sent to NBS for metallographic examination confirmed the G.E. observations (figure 1). Nevertheless, it was decided to subject the specimens to electron probe microanalysis. Figure 2 shows scanning images for Fe, Cr and Ni. It is apparent that none of these three elements are homogeneously distributed. Oxygen was not studied since the O-K line lies between the Cr-La and Cr-L lines and cannot be separated from either. Random point counts (avoiding Cr-rich areas insofar as possible) listed in table 1 show swings of as much as 20% of the Ni and Fe in ferrite depending on position. Therefore, it must be concluded that this attempt to produce stainless steel of high homogeneity by powder metallurgical techniques has failed. However, However, signal success has been achieved in preparing homogeneous W-20% Mo by a similar powder process thus indicating the usefulness of this approach. It seems clear that the failure of the stainless steel preparation rests solely on the impossibility of obtaining suitable Cr starting powder. 50 REFERENCE [1] P. Duncumb and D. A. Melford, Quantitative Applications of Ultra-soft X-ray Microanalysis in Metallurgical Problems, Tube Investments Res. Lab. Tech. Rept. No. 195, 27 pp (1965). |