Microscopic Study of Strain in Metals. By F. Rogers. 17 



-andulations in soft irons — Swedish and Low Moor — was found to 

 be from three to five times the average breadth of crystal grains in 

 the specimen. The author has since found a similar relation 

 between the wave-length and size of grains in both normal and 

 overheated samples of steels of the series referred to above, upon 

 loading them statically in tension. Further, it is found that the 

 ruffling of the surface, observed in fatigued samples of the normal 

 class, has also a similar relation to the structure. 



The crests follow very irregular lines on the observed surfaces ; 

 this and the variability of the wave-length point to the dependence 

 of the form, dimensions, and indeed existence, of the undulations 

 upon the microscopical heterogeneity of the material. 



The properties of these undulations, and the observation that 

 the slips in the normal samples when fatigued are numerous — 

 though not so numerous, it should be remarked, as in similar 

 specimens severely overstrained statically — show that there is a 

 tendency for the inelastic and injurious strain to be much more 

 minutely subdivided and uniformly distributed throughout the mass 

 of the normal than of the overheated steels. This helps to explain 

 why overheated steels are less enduring than normal steels under 

 alternating stresses — a fact which is now established beyond doubt 

 by Stead's experiments and the author's. 



When fatigue has proceeded to a late stage, it is possible to 

 see a fine crack in ductile metals without the aid of the Microscope ; 

 and in the last few reversals before rupture the crack is usually 

 widely open. But the cracks may have developed to such an 

 extent as to weaken the piece dangerously some time before they 

 are visible to the naked eye, as the following results show. The 

 weakness of such a piece would at once be evident in a static 

 or dynamic test. The primary object of the series of experiments 

 was to determine the probable effect of annealing upon the 

 further endurance of partially fatigued metal, and showed that 

 even at a comparatively early period of the fatiguing, the cracks 

 may have made so much headway that annealing is practically 

 useless. Upon further fatiguing to rupture a specimen which is in 

 this condition, the progress which the cracks had made up to 

 the time of the intermediate annealing is clearly mapped in tint 

 upon the final fracture. The shorter the final stage of the en- 

 durance, other things equal, the greater the heat-tint markings 

 upon the fracture. The examples given in the table on the follow- 

 ing page illustrate this. 



Plate IV. fig. 5 shows the fracture of specimen No. 6 at about 

 17^ diam. The dark segmental areas are heat-tint marks. These 

 results probably show the cause of difference of opinion existing 

 as to the value of periodical annealing of machine and other 

 parts which are submitted to heavy service. Clearly, annealing 



Feb. 20tK 1907 C 



