154 



METALLURGICAL MICROSCOPES 



ascribed to the S2)litting up of the solid solution, the iron carbide being 

 precipitated in a fine state of division throughout the solvent iron. 

 That this explanation is correct in its outline is more than probable, and 

 when the tempering has proceeded to something like completion the 

 presence of the carbide is easily detected. The early stages of the 

 decomposition are practically incapable of observation v^ith the present 

 microscopic means which are available, but this is the portion of the 

 process which is of the greatest interest and importance. It is almost 

 impossible at the present time to say at what temperature the decom- 

 position of the solid solution commences or how it proceeds. In many 

 carbon steels the maximum stress, which in the ordinary way is 

 supposed to decrease when the tempering temperature increases, 

 actually increases at tempering temperatures near to 400° C. Typical 

 results are : — 



Table i. 



It is quite likely that if the constitutional changes going on within 

 the Steel could be examined microscopically, and a better idea of these 

 changes formulated, the explanation of the peculiar happenings would 

 be found. This would require a very high power — something probably 

 of the order of ten times as great as the present powers. 



Similiar problems arise in connection with the tempering of the 

 alloy steels. In the nickel chromium steels, for instance, a property 

 known as temper brittleness is shown by steels which cool slowly from 

 or through a certain range of temperature, e.g., 425° C. to 550° C, 

 In all other respects the mechanical properties of the steels are the 

 same whether cooled quickly or slowly. The effect of different methods 

 of cooling from this temperature upon the toughness is shown in 

 Table 2. 



It is inconceivable that there is no difference at all in the con- 

 stitution of the two steels, but the present microscopic methods fail 

 entirely to detect the difference (see Figs. 1 and 2). It is possible that 

 higher powers would make the detection possible. The same powers 

 might also give an explanation of the peculiar impact values which are 

 obtained from the alloy steels by tempering at comparatively low 

 temperature, e.g., 150° to 350° C. The accompanying curve (Fig. 3) 

 shows the values which are customarily obtained, and in addition shows 



