Ductile Materials under Combined Stress. 7 1 



is no evidence of drawing out, and the fracture is practically 

 normal to the axis of the specimen. When ductile speci- 

 mens are broken in torsion, the surfaces of rupture are normal 

 to the axis ; but in the case of cast iron the surface of rupture 

 takes a curious form, one of the edges of the fracture of a 

 circular cylindrical specimen being a helix whose angle is 

 about 45°. Some steels of a treacherous nature exhibit a 

 fracture partaking of the two types, part of it being normal to 

 the axis, but part extending along the length of the specimen 

 and having the helical edge. Under compression ductile 

 materials ultimately give way by lateral now ; cast-iron spe- 

 cimens exhibit rupture-surfaces inclined to the axis at 

 about 30°. 



It will be noticed that in all cases the ductile materials 

 exhibit rupture-surfaces coincident to a large extent with the 

 surfaces of maximum shearing- stress, suggesting that the final 

 action in such cases partakes of the nature of a viscous flow. 

 Cast iron, however, under tension and torsion fractures along 

 the surfaces across which the tension or elongation is greatest; 

 in compression it appears to fail by a shearing action, in- 

 fluenced perhaps by the friction of the crystals. 



These illustrations of modes of fracture tend to show that a 

 sharp line of demarcation may be drawn between the behaviour 

 under stress of ductile and other materials, and this in spite 

 of the gradual transition from one stage to the other in 

 manufacture and in use. 



Owing to the increasing tendency to substitute steel for 

 cast iron in structures and machines, the ductile materials 

 were selected as the subject of experiment; steel of four grades, 

 copper, and brass have been experimented upon. 



o. The Yield-point in Tension Tests. — Upon observing the 

 operation of testing more carefully and taking measurements 

 of the extension at frequent intervals, we notice that under 

 the lower loads the specimen stretches so very slightly as 

 hardly to affect the beam of the testing-machine. This con- 

 tinues until we reach a point called the yield-point, marked 

 by a large extension of the specimen for a slight increase of 

 load, the lever of the testing-machine in consequence moving 

 considerably. At the same time the scale, if there is any, on 

 the specimen peels off, starting by cracks inclined at about 45° 

 to the generating-lines, separating the scale into squares; or if 

 the specimen be machined, certain characteristic changes take 

 place upon its surface. After some time the stretching under 

 this load ceases and the load on the specimen can be increased, 

 the elongation also increasing until rupture takes place. 

 The loads and the measurements of the extension they pro- 



