74 Mr. P. W. Bridgman on Breaking Tests under 



will occur for a material like Bessemer -teel at an intern.il 

 pressure about 4/5 of the tensile strength. 



The tact is, however, that for the ordinary materials of 

 engineering practice rupture begins at the outside and runs 

 in towards the centre. This rupture may take place either 

 by tearing apart of the metal in an axial plane, or else by 

 slip on a shear plane, the fracture in this latter event 

 running in toward the centre approximately as an equi- 

 angular spiral. 



Most of these bursting tests were made on drawn bars, 

 which were pierced with the hole first and then turned off 

 slightly on the outside so as to be concentric with the hole. 

 Of course the most obvious suggestion with regard to this 

 rupture from the outside is that there were flaws in the 

 outer skin of the drawn bar penetrating more or less deeply 

 into the interior, and that the rupture started from one of 

 these flaws. In order to prove definitely that this is not the 

 case, the following test was made. A bar of Bessemer steel, 

 4^ in. diameter, was turned down to 4 in. so as to remove the 

 outer layer. From this 4 in. bar a number of rings, 4 in. in 

 diameter and 1/8 in. square in section, were turned at regular 

 axial intervals of about 1 in. From the bar left after cutting 

 off the rings, a cylinder for testing was made in the usual 

 manner about 3| in. o. d. and ^ in. i. d. and 8 in. long. From 

 the other end of the same bar a similar ^ct of rings and a similar 

 cylinder were turned, only smaller, the cylinder being 2 in. 

 o. d. and ^ in. i. d. The rings were then tested to rupture 

 on an expanding mandrel. No trace of flaw in the steel 

 was found ; the rings expanded between 10 and 15 per 

 cent, before rupture, and the location of the fracture was 

 haphazard, showing no longitudinal vein of special weakness 

 in the steel. The cylinders were then tested to rupture in 

 the usual manner. For this purpose, where the cylinders 

 are made of very ductile material, it was found convenient 

 to fill the cylinders with lead instead oi with a true liquid, 

 since it is easier io keep the lead from leaking after the 

 cylinder has begun to stretch. The had transmits these 

 high pressures nearly hydrostatically. Pressure was pro- 

 duced by a hardened steel piston forced against the lead by 

 the ram of an hydraulic press. A cup-shaped washer of 



Bessemer Bteel prevented the Lead from leaking past the 

 piston. Because of the very great Middling it was 

 necessary to make several strokes of the piston before 

 rupture was produced. 



The results of these two tests with the two cylinders 

 mentioned above are Bhown in the photographs. The larger 



