FRACTURE OF METALS UXDER REPEATED ALTERXATIOXS OF STRESS. 249 
They also help to explain why it is that a piece that has been subjected to many 
reversals shows no apparent loss of strength or plasticity when subjected to an 
ordinary tensile test. So long as the reversals have not yet reached the stage ot 
producing cracks, it is not to be expected that such a test will detect the 
deterioration which has occurred. The material will still yield by slipping much as 
at first, and neither its plasticity nor its strength need show much change. 
Interesting points suggest themselves which require further investigation. It is 
well known that when a plastic metal such as iron is strained sufficiently to take 
j)ermanent “ set,” it suffers a temporary loss of elasticity, which is recoverable by 
lapse of time, the recovery of which, as Muir has shown, may be enormously 
accelerated by warming the piece to such a temperature as 100° This may be 
ascribed to a gradual healing action which restores the resistance to sliding on the 
planes of slip after they have been weakened by the first severe strain. The first 
strain makes subsequent slipping easier, for a time, but when the material has a long 
enough rest recovery ensues. Probably enough a similar recovery would occur if 
during the application of reversals of stress long intervals of rest were allowed, and 
still more if during these intervals the temperature of the piece were raised. It may 
be conjectured that such treatment would arrest the destructive process of fatigue in 
its earlier stages, and give the material a new lease of life. The damage which 
alternating stresses produce probably depends not only on the amount of the stress 
and the number of alternations it suffers, liut on the rapidity with which the 
alternations follow one another,! and on the continuity or otherwise of the alternating 
action. 
Description of Plates. 
PLATE 7. 
Figs. 3-6. Fatigue of Swedish iron by reversals of a stress of 14‘3 tons per sq. inch. 
Magnification 150 diameters. 
Fig. 3 after 5,000 reversals. 
Fig. 4 after 40,000 reversals. 
Fig. 5 after 60,000 reversals. 
Fig. 6 after 70,000 reversals, followed by re-})olishing and re-etching. 
* Muir, “ On the Recovery of Iron from Overstrain,” ‘Phil. Trans.,’ A, vol. 193, p. 1. 
t On this point, the experiments of Osborne Reynolds and J. H. Smith (‘Proc. Roy. Soc.,’ vol. 70, 
1902, p. 44) have already shown that rapid reversals are more destructive than less rapid reversals, in the 
sense that to make fractures result from an equal number of both, the maximum stress must be higher 
when the number of reversals per minute is less. 
VOL, CC.—A. 
■ ) 
K 
