HIGH PRESSURE BRIDGMAET. 187 



as high as 300,000 pounds per square inch in a solid piece of steel, 

 but it is another matter to maintain such a pressure in a liquid and 

 prevent all leaking of the latter from the container. 



The most essential part of the preliminary work was to design 

 a packing that would keep the vessel in which the pressures were to 

 be produced absolutely tight, and prevent the liquid from leaking 

 from it. The feature of the form of packing finally designed is that 

 it is made tighter and tighter by the pressure in the vessel itself; 

 the greater the pressure in the vessel, the less can the liquid leak. 



The second part of the preliminary work consisted in finding 

 what limits of pressure steel apparatus will support, steel being 

 selected as the best material of which to make the pressure im- 

 plements. In all the experiments the pressures were produced by 

 pushing a piston, by means of an hydraulic press, into a cylinder 

 containing the liquid. In the piston the strain is one of compres- 

 sion, while in the cylinder it is one of bursting, or tension. 



It was found by experiment thai the best material for the piston 

 was glass-hard steel. The compression that a piece of glass-hard 

 steel will support when it is held rigidly so that it will not bend is 

 surprisingly large. Several grades of steel were found that would 

 support a compression of 600,000 pounds per square inch, and one 

 grade supported as high as 750,000 pounds per square inch. 



The strength of the steel cylinders was also a factor which had 

 to be settled by experiment, since it was found that no theory of 

 the strength of a cylinder is of any value for very high pressure. 

 All ordinary theories predict that no cylinder can be stressed to 

 more than the tensile strength of the steel, no matter how thick are 

 its walls. A few rough experiments showed the actual pressure 

 that a cylinder can support to be much in excess of that predicted 

 by the ordinary theory; this is on account of the fact that when 

 the pressure reaches a certain value the inner layers do not break 

 but stretch, and thus allow the outer layers to assume some of their 

 share of the load. It was found that the most efficient way to make 

 a cylinder support a high pressure was first to stretch it on the 

 inside by applying a much higher pressure than it was intended to 

 maintain in practice, and then to machine it to its final diameter. 

 A cylinder treated in this way is in a state of internal strain, exactly 

 as is a gun which has hoops shrunk on it from the outside, the 

 tension in the hoops inducing initial compression in the interior of 

 the gun. When pressure is produced in such a gun, it removes the 

 compression from the inner layers of the material and the tension 

 of the outer layers is increased. But it has been shown that the 

 tension in the inner layers increases more rapidly than that in the 

 outer, and it can be seen, therefore, that, in time, the increasing 



