44 



MR. LEONARD BAIRSTOW ON THE ELASTIC LIMITS OF 



had also become constant and small, but not zero. The rate of extension during the 

 day was almost always greater than that during the night, and this effect is supposed 

 to be due to change of temperature as the magnitude of the change from day-rate to 

 night-rate was greatest in frosty weather. The mean for the whole curve is regular, 

 the earlier part of the curve having taken about 11 days for completion, involving a 

 fairly good daily average. 



Increase in the range of stress produced a greater rate of extension after a 

 preliminary decrease due to the larger compression load. Equal stresses have been 

 found, as previously stated, not to produce any appreciable permanent extension and, 

 as the method tends to equality of loads, it would be expected that the rate of 

 extension could not become very great in this case. 



Oh] 



II 



uc 

 u 



Q. 



o 10,000 20,000 



NUMBER or REPETITIONS or STRESS 



O IO ZO 



STRESS. TONS PL* SQ IN 



Fig. 3. 



Fig. 3. The maximum stress applied, i.e. + 23'2 tons per sq. inch, was still below the 

 ordinary yield-point for the material, and for the first 2000 repetitions very little 

 appeared by the extensometer. Shortly afterwards somewhat rapid permanent 

 extension began, fig. 3, a, which corresponded to a slow yield, and at the same time a 

 hysteresis loop, fig. 3, c, made its appearance. The latter reached its maximum width 

 in about 7000 repetitions and then remained unaltered for a further 8000 repetitions. 

 During this latter time the rate of extension was gradually becoming less, until 

 finally it was very small. A slightly decreased range would have allowed this part of 

 the curve to become horizontal and the conditions would have been stable. Some 

 " cyclical permanent set " might then have remained temporarily, but there is every 

 reason to believe that it would have disappeared by recovery if sufficient time were 

 allowed. 



