. TRANSACTIONS OF THE SECTIONS. 113 



On the Constmction of Iron Vessels exposed to sei'^ere strain. 

 By William Fairbairn, C.E., F.R.S. 



In the construction of vessels, such as boilers, pipes, &c., exposed to severe inter- 

 nal pressure, it is desirable to obtain some knowledge of the strength and condition 

 of the material used, and some fixed rules calculated to enable us to judge vf\t\\ ac- 

 curacy as to the disposition of the parts in order to apply the greatest strength in the 

 direction of the greatest strain, and, in fact, so to dispose of the material in order 

 that every part of the vessel shall balance itself in its power of resistance when sub- 

 jected to uniform pressure. 



To attain these objects, the author gave the results of his experiments on the re- 

 sistance of malleable iron plates first announced to the British Association, and sub- 

 sequently published in the Transactions of the Royal Society. These experiments 

 were originally undertaken to determine the strength of metal plates, beams, and 

 angle-iron, as applied to ship-building ; and they have since been continued, from 

 time to lime, for the equally important purpose of improving the construction of 

 malleable iron bridges, boilers, and other vessels, such as caissons and sheet-iron 

 pipes, which are now coming into more general use for pump trees and other arti- 

 cles connected with mining. 



In order to acquire satisfactory data on the strength of the material employed, a 

 variety of plateS from Low Moor, Staffordshire, and other parts, were submitted to 

 direct experiment ; first, by tearing them asunder in the direction of the fibre ; and 

 secondly, across it. The tensile strength per square inch was ascertained to be as 

 follows : — 



In the direction of the fibre. Across the fibre. 

 Yorkshire plates . . = 24*26 26-93 



Derbyshire plates = 2].-68 18-65 



Shropshire plates = 22-82 20-00 



Staffordshire plates = 19-56 21-01 



Mean in tons 22-16 Tons 22-29 



From this it will be observed that there is no difference in the strength of iron 

 plates, whether torn in the direction of the fibre or against it ; and this uniformity of 

 strength probably arises from the superior manner in which that article is now 

 manufactured. 



The experiments would however be imperfect as regards construction, if they had 

 not been extended to the process of riveting; and on this point our information has 

 been of the most meagre description. Until of late years many of our numerous 

 constructions have been conducted under the impression that the riveted joint was 

 not only strong, but absolutely stronger than the plate itself; whereas more than 

 one-third of the strength is lost by that process. 



To prove the fallacy of these views, it was ascertained by experiment that the 

 strength of iron plates, as compared with their riveted joints, was not only weak- 

 ened to the extent of the quantity of metal punched out to receive the rivets, but 

 that in the following ratios, viz. as 1000 : 700 in the double-riveted joint, and 

 100 : 560 in the single-riveted joint. 



From the above facts practical formulae have been deduced to show, that the 

 maximum resistance of single-riveted plates does not exceed 27,000 lbs. to the square 

 inch ; but taking into account the crossing of the joints and other circumstances 

 peculiar to sound construction, 34,000 lbs. or 15 tons per square inch has been found 

 to be the maximum strength of riveted plates, such as those used for boilers and 

 similar constructions. 



In conclusion, attention was directed to several important improvements in con- 

 nection with the construction of steam-boilers, by the introduction of gussets, to 

 strengthen the flat ends and retain them in shape. After noticing that all boilers 

 should be of the cylindrical form, Mr. Fairbairn observed that when flat ends are 

 used they should be composed of plates one-half thicker than those which form the 

 circumference. The flues, if two in number, to be of the same thickness as the ex- 

 terior shell, and the flat ends to be carefully stayed with gussets of triangular plates 

 and angle-iron, connecting them with the circumference and the ends. 



1851. 8 



