MECHANICAL TESTING OF IRON AND STEEL. 221 



Since this involves but very few extra measurements and 

 very little time, in fact they can all be made and the results 

 worked out while the next bar is being- fitted into the 

 machine, it is a pity they are not systematically recorded 

 in all tests. 



Work of Fracture. Professor Kennedy pointed out 

 in a discussion at the Institute of Civil Engineers {Pro- 

 ceedings, vol. Ixix., p. 30), that the amount of work spent 

 in fracturing any bar, expressed in inch tons per cubic inch 

 of the bar, was a most valuable indication as to the power 

 of any given material to resist sudden shocks or blows. 

 Whenever an autographic stress-strain diagram has been 

 drawn during the fracture of a bar it is an easy matter 

 to calculate the work of fracture, since it is proportional to 

 the area enclosed between the stress-strain curve and the 

 axis on which the strains are set oft, and its numerical value 

 is known directly we know this area in square inches and the 

 scales of stresses and strains ; it can also be similarly deter- 

 mined when the curve has been plotted afterwards from the 

 observed strains recorded during a test. But as both these 

 operations involve a good deal of labour and time, it has 

 generally been assumed that the determination of this factor 

 wasbeyond the scope of ordinary testing. Professor Kennedy, 

 however, in this discussion gave a formula for obtaining the 

 work done in fracture, involving only a very short arithmet- 

 ical calculation ; he pointed out that all the data required in 

 his formula were known from the observations always made 

 during- an ordinary test. The formula is that the work done 



wx 



(r± 2) 



in inch tons = — — \ where iv is the maximum load 





carried in tons per square inch, x is the final extension in 

 inches and r the ratio of the Limit of Elasticity to the Maxi- 

 mum load. By Limit is here meant the more scientifically 

 termed Yield Point or break-down point, which is always 

 readily enough indicated during the test by the steelyard drop 

 and therefore involves no delicate instrument for measure- 

 ment. All the three factors of the formula are known directly 

 the test is complete, and with the aid of a slide rule the simple 

 calculation required is made in a few moments without causing 



