The Rupture of Steel by Longitudinal Stress. 243 



Schiaparelli (M. E% G. V.). Considerazioni sul Moto Rotatprio del 

 Pianeta Venere. 8vo. Milano 1890. With Four other Excerpts 

 in 8vo. The Author. 



Two Folio Volumes containing MS. Correspondence on Terrestrial 

 Magnetism, between the Rev. Humphry Lloyd and Sir E. Sabine 

 and others. 18331878. Mrs. Lloyd. 



rt The Rupture of Steel by Longitudinal Stress." By CHARLES 

 A. CARUS- WILSON. Communicated by Professor G. H. 

 DARWIN, F.R.S. Received March 10, Read March 27, 



38 HO. 



[PLATKS 2, 3.] 



In a paper read before the Royal Society on June 16, 1881, 

 Professor G. H. Darwin stated : " It is difficult to conceive a^Tiy mode 

 in which an elastic solid can rupture except by shearing, and hem e 

 it appears that the greatest shearing stress is a proper measure of 

 the tendency to break" ('Phil. Trans.,' 1882, p. 99). 



In this paper, I have recorded the results of some experiments 

 made with a view to throwing light on the question raised by 

 Professor Darwin. 



The experiments were conducted in the mechanical laboratory at 

 the Royal Indian Engineering College, Cooper's Hill, the machine 

 used being a hundred-ton single lever hydraulic testing machine by 

 Messrs. Buckton & Co., of Leeds. This machine is fully described in 

 Unwiu's ' Testing of Materials of Construction,' p. 133. 



All tension experiments were performed on circular specimens 

 fitted with a screw thread at each end, on to which were screwed 

 steel nuts resting in spherical seatings to ensure directness of pull. 



The shearing experiments were conducted on specimens screwed 

 throughout their entire length into three steel blocks,. and tested in 

 double shear, i.e., the two outside blocks were pulled in the opposite 

 direction to the inside block perpendicularly to the axis of the 

 specimen, and the bar sheared in two places at once. 

 . The idea of rupture necessarily implies an overcoming of resist- 

 ance. If the power of steel to resist rupture under tension were a 

 constant quantity, the conception of rupture would be simple, for 

 we should only have to increase the stress up to the required amount, 

 and the bar would break ; but the resistance to rupture appears to 

 be a function of the amount of flow that has taken place (see experi- 

 ments described in Table II). If we take two axes, OX, OY, to re- 

 i present respectively the elongation per unit of length, and the stress 



r uuit of area of transverse section of elongated bar, we obtain the 



pe 



