Variations in the Average Size and Quantity of Nonmetallic Inclusions during Oxidation and Deoxidation of Steel G. K. BANDYOPADHYAY, H. S. RAY, AND R.K.GUPTA An investigation has been c a r r i e d out to study the effects of the degree of oxidation or deoxida- tion of liquid steel on the average size,number, and quantity of nonmetallic inclusions. During deoxidation the variations are strongly influenced also by the nature of the deoxidizing elements and may, therefore, be used to study the individual characteristics of the elements. Some cal- culations of the oxygen and inclusion contents of the bath are made on the simple assumption that all inclusions are spherical and have the same radius at any given instant. It is shown that such calculations may be used to obtain reasonably reliable estimates. T H E commercially available steels usually contain a variety of nonmetallic inclusions such as oxides, ni- trides, or sulfides. The oxides are usually the most predominant and, therefore, have received a greater amount of attention in the literature. These inclusions mostly come from the stable oxide particles which are formed and trapped during the deoxidation of molten steel. The inclusions are known to have significant in- fluence on the mechanical properties of steel.1 In some instances, the inclusions may well serve to enhance se- lected properties, while in some others they may be extremely detrimental. The prevailing effect is deter- minedby the size, shape, quantity, and the distribu- tion of the inclusions. During oxidation or deoxidation reactions, the changes in the oxygen level of the bath and the nature of the oxidizing or reducing agents should directly af- fect the variations in some secondary parameters such as the average size, the average number per unit volume, and the volume fraction of the inclusions. Some recent investigations have been directed at the kinetics of the formation and the growth of the deoxida- tion products,2-4 and the effective removal of these nonmetallic inclusions from the m e t a l b a t h . s'6 In the literature, few systematic studies have been reported on the variation of the various secondaryparameters during oxidation and deoxidation of steel. Such studies would not only describe the variation of the oxygen level and the nature of the deoxidizer but may serve as important guidelines for assessing the quality of steel viewed as a metal-oxide composite. In the present investigation, the variations in sev- eral secondaryparameters have been studied system- atically during the oxidation of steel by direct oxygen injection and deoxidation usingvarious deoxidizers such as A1, Si, Fe-Si (79 pct Si), Fe-Mn (75 pct Mn), Fe-Cr (75 pct Cr), and graphite. The reactions have been carried out in commercially available graphite crucibles at 1550°-~ 10°C using an induction furnace. G. K. BANDYOPADHYAY and H. S. RAY are Graduate Student and Assistant Professor, respectively, Department of MetallurgicalEn- gineering, Indian Institute of Technology, Kanpur (U.P.), India. R. K. GUPTA is Metallurgist, Indian Smelting and Refining Co. Ltd., Bhandup, Bombay, India. Manuscript submitted February 16, 1970. EXPERIMENTAL PROCEDURE About 500 g of metal of variable carbon content was melted in the induction furnace and oxidized for a few minutes at 1550° ± 10°C by direct injection of oxygen on the metal surface using an alloy steel tube. During one particular oxidation process samples were col- lected at regular intervals by quickly pouring out 10 to 15 g of m e t a l into refractory molds. For the other h e a t s , only the ultimate oxidized steel was sampled. Next, a weighed quantity of crushed deoxidizer (-50 +100 mesh) was a d d e d w r a p p e d in a piece of paper and the melt was stirred with an alloy steel rod. The at- tack on the alloy steel rod was negligible. The deox- idizers used were A1, Si, Fe-Si (79 pct Si), Fe-Mn (75 pct Mn), F e - C r (75 pct Cr), and graphite powder. The quantity of deoxidizer was fixed at 0.035 pct by weight except in the case of silicon where a much larger amount was added (2.5 pct). It is known that below 2.0 pct the deoxidizing power of silicon is very limited because a good part of it is consumed by atmospheric oxidation. The graphite powder was very fine (-100 mesh). It was injected into the melt with the help of a specially designed graphite injector, with consumable graphite nozzle, under 5 psi pressure of argon gas. The total amount of graphite injected in 6 min amounted to about 0.021 pct. Having added and stirred the deoxidizers, steel sam- ples were poured out into refractory molds at different intervals. The solidification of the samples in the molds was rapid and the possibility of the segregation of inclusions due to flotation could be ignored. All samples were subsequently prepared and polished for microscopic examination at room temperatures. The size of inclusions was measured by using an eyepiece engraved with a centimeter scale and an ob- jective of suitable magnification (usually 10X). To get the true statistical average more than forty random measurements were done on the sample surface. The averagenumber of inclusions per unit area on the surface, was obtained from direct counting on at least ten different locations. The number density of inclusions, defined as number of inclusions per unit volume, was calculated from the above measurement using a mathematical formula discussed subsequently. The area density of inclusions reported here is a METALI_URGICAI. TRANSA('TIONS VOLUME 2, JANUARY 1971 239