30 NON-ELASTIC DEFORMATION. Art. 7 



ally plastic quality the bar may in this way finally be caused to 

 break with a fracture resembling that of hard steel, with com- 

 paratively little total extension or contraction of section at the 

 fracture, and under a total load much greater than that which 

 could be applied in an ordinary test. ' ' 



27. Hardening Effects of Non-elastic Deformation. After 

 a piece of iron or steel has undergone a permanent set, it is 

 hardened, that is, it is less ductile and its ultimate strength is 

 increased. This hardening is of great practical importance, and 

 may be considered a benefit as in cold-rolled or cold-drawn shaft- 

 ing and in wire for certain puposes, or an injury as in wire for 

 other purposes, and in steel hammered or bent cold, or subjected 

 to other severe treatment like punching and shearing. A very 

 narrow strip of metal contigious to a punched hole or a sheared 

 edge is hardened so much as to make it brittle. Under stress, 

 this hardened portion is not capable of deforming as much as the 

 other metal, and therefore causes a non-uniform distribution of 

 stress and a reduction of ultimate strength. The softer and thin- 

 ner the metal, the less the injury ; for the best class of work, it is 

 the practice to remove this hardened part, in medium steel, by 

 reaming out punched holes and planing off sheared edges. 



All of the effects of non-elastic stress may be much modified 

 or entirely obliterated by annealing, that is, by heating to red- 

 ness and cooling slowly. Wire is annealed for certain purposes 

 which require a soft wire and, in this manner, any piece whose 

 uniformity has been disturbed by cold working or partial 

 heating is restored. 



28. Fatigue of Iron and Steel. If in place of a few appli- 

 cations of increasing stresses, as explained in Art. 26, many 

 applications of the same stress are made, the elastic limit is raised 

 if the stresses do not approach too closely the original ultimate 

 strength, and will be raised considerably higher than the applied 

 stresses if these are not very much above the original elastic 

 limit 1 . With continuously repeated stresses above the rising 

 yield point, rupture will finally occur at a stress much below the 

 original ultimate strength, and with a fracture like that of a 

 brittle material or like that of a piece of steel with an artificially 



! This was investigated bf Banschinger ; for the results of his experi- 

 ment's, see Unwin's T*\e Testing of the Materials of Construction, Chap- 

 ter XII. 



