ee eS 
COMPLEX STRESS DISTRIBUTIONS IN ENGINEERING MATERIALS. 195 
importance of a flaw such as a crack, a non-metallic enclosure, or even a 
tool-mark, is relatively very great, and depends upon the composition 
generally, increasing, for example, with the carbon content. Heat treat- 
ment may increase or decrease the relative importance of such flaws 
according to the kind of treatment and the previous condition of the steel. 
Apart from flaws of the kinds mentioned, steel in the rolled or forged 
condition occasionally happens to be weak dynamically. For the present 
purpose it appears necessary to consider the effects of heat treatments 
upon steels in the rolled or forged condition which are not weak from 
either of these causes. 
There are then three main classes of treatments to consider :— 
(1) Overheating —This in general diminishes the endurance under 
alternating stress (62 and 87). When extreme it merges into ‘ burning,’ 
from which it is distinguished technically. Shght overheating, on the 
other hand, is the same thing as some of the processes which are called 
annealing. 
(2) Reheating through the critical range is, in general, capable per se of 
bringing the endurance to a normal high value, or of leaving it undisturbed, 
according to the state of the steel before the treatment (Nos. 62, 85, 86, 
and 87). The following factors tend to make the effects of such reheating 
approximate more and more to those of overheating: (a) the more the 
temperature exceeds the upper limit of the critical range ; (b) the greater 
the duration of heating above the lower limit of the critical range; (c) the 
slower the cooling through the critical range (62). 
(3) The speed of cooling through the critical range has in any event a 
most profound influence upon the endurance under alternating stress. 
Generally speaking, it appears that the more rapid this cooling the greater 
is this endurance (62, 33). 
As to the effect of Cold Work upon endurance of alternating stress, 
there are no data available. It is well known to manufacturers that it 
increases the endurance greatly in some cases—for example, wire. This 
fact also explains partly why in some published experiments (e.g., 62) 
“annealing ’ diminished the endurance. The bars from which the tests 
were cut were of small section, and therefore they were somewhat cold 
worked, and also relatively rapidly cooled, in manufacture. They were 
much more slowly cooled in some of the experimental annealings. 
The effect of annealing after a metal has withstood large numbers of 
alternations is also one which can only be answered when many practical 
particulars of the metal are known. In (47) no effect was found. In (62) 
it was clearly proved why no effects could be obtained from annealing after 
a certain stage of the fatigue had been passed. At a comparatively early 
stage minute incipient cracks are sufficiently open to contain air. Hence 
the faces oxidise, effectually ending any possibility of reunion. 
Note on Microscopic Effects of Alternating Stress. 
By Dr. F. Rogers. 
This has been exhaustively elaborated in a very few papers. The main 
conclusion is to show that cracks form by the development of repeated 
cleavage, seen as slip-bands. This was done in (27) for iron, and in 
(62, 63, and 82) for steel. In (63) and (82) the influence of the constituents 
is noted, and in particular the avoidance of the harder carbon containin g 
0 2 
