102 BELL SYSTEM TECHNICAL JOURNAL 



troostite occurs which was described as flocculent border type for lack 

 of a better designation. It seemed to be largely ferrite and appeared 

 to be the means by which the excess constituent (in this case ferrite) 

 appeared at the grain boundaries. 



If a small specimen of the same steel is heated to a high tempera- 

 ture and quenched in oil or water depending on the circumstances of 

 the experiment, a structure results which may be largely martensite 

 needles with scattered particles of troostite. Sometimes relatively 

 large areas on the prepared surface of the specimen may be almost 

 entirely of the constituent troostite. As is well known, this condition 

 is controlled by the rate of cooling in the quenching operation. The 

 type of troostite found in uniformly heated and quenched specimens 

 was defined as nodular troostite. This paper deals further with this 

 particular constituent of hardened steel. 



Since these early experiments in which high power metallography 

 was first applied with success to the structures of hardened steel, there 

 have been many improvements in technique. These improvements 

 have resulted in a much higher order of resolution and it is the object 

 of this paper to review the past work in the light of the improved meth- 

 ods now available and to present some new results. 



To quote from the Franklin Institute paper: 



"... These nodules develop from innumerable nuclei throughout 

 the austenite and martensite matrix. . . . The nuclei increase in 

 number and the developing nodules become larger and larger. Irregu- 

 larities in growth due to interference of nodules occur as the growing 

 particles increase in number and size until finally the whole mass seems 

 to be composed of nodules, some spherical in shape but many deformed 

 due to mutual interference and to irregularities in growth. A selec- 

 tivity or preference in crystal habit probably prevails for crystallo- 

 graphic planes since spines, branches, and interconnected crystallites 

 may be found occasionally. In reality these are poorly formed nodules, 

 growth in some one or more directions having been arrested." 



" It is quite evident that if the entire mass of the metal passes through 

 the nodular troostitic stage, this constituent must contain carbide or 

 carbon in some form." 



"When one of the globular-shaped crystal masses which has developed 

 under favorable conditions of growth is sectioned in such a way as to 

 divide the mass along a plane passing through the center; the nucleus 

 is found at the center and fan-shaped grains extending from the center 

 toward the outside. When freshly formed and under the highest pow- 

 ers of the microscope these radial grains have all of the appearance of a 

 solid solution. The nodule must contain carbon in some form, as 

 stratification soon takes place." 



