106 BELL SYSTEM TECHNICAL JOURNAL 



Professor Honda believes that martensite forms first and troostite 

 develops secondly, replacing the martensite. '^ In his discussion of the 

 subject he appears to deal with the ultimate nature and composition of 

 the constituents and not with their outward form. 



A number of typical illustrations are included to show in detail the 

 structure of troostite. For these experiments a high grade tool steel 

 of about 0.90 per cent carbon was used. Small specimens weighing 

 about 10 grams were suitably heated in a vacuum furnace to a high 

 temperature and quenched in ice and brine solution. The hard- 

 ness of the specimens was quite uniform and averaged C-65 on the 

 Rockwell scale. From a study of the photographs it is quite apparent 

 that in a specimen of the kind, we may have not only the constituents 

 martensite and troostite, but in the troostite also the constituent pearl- 

 ite in the form of fan-shaped grains. From the work of Mathews,^ 

 Bain,^ Enlund,'" and others, it is also apparent that some retained 

 austenite may be present. Hardened steel, therefore, is a complex 

 structural aggregate at best. 



Conclusions 



Nodular troostite appears to be an aggregate of ferrite and carbide, 

 and in the very early stages of formation its structure is on the border 

 of present methods of resolution. The condition of the ferrite and car- 

 bide in relation to each other is not stable; they tend to stratify form- 

 ing pearlite. 



Troostitic nodules grow about a nucleus which may be an inclusion, 

 a void, a corner in a grain boundary or some other detail of structure. 

 The nodules contain fan-shaped radial grains. 



The development of troostite results in a reorientation of the ferrite, 

 seemingly without particular reference to the old austenitic crystal- 

 lographic planes. Martensite does follow the old system of austenitic 

 planes. 



The small fan-shaped grains in nodular troostite may persist as small 

 grains or they may undergo grain growth by union. It is a matter 

 seemingly dependent upon the thermal treatment of the specimen. 



^ Honda, "Is the Direct Change from Austenite to Troostite Theoretically Possi- 

 ble? " Journal British Iron and Steel Institute, 1926, Vol. GXIY, No. 2. 



* Mathews, "Austenite and Austenitic Steels," Trans. American histitute of Mining 

 and Metallurgical Engineers, Vol. LXXI, 1925; "Retained Austenite," British Iron 

 and Steel Institute, 1925, No. 11, Vol. CXII. 



'Bain, "The Persistence of Austenite at Elevated Temperatures," Tratis. 

 American Society for Steel Treating, \'ol. VIII, 1925. 



1° Enlund, "On the Structure of Quenched Carbon Steels," Journal British Iron 

 and Steel Institute, 1925, Vol. CXI, No. 1. 



