AUTOGRAPHIC DIAGRAMS. 241 



having the same dimensions. In these torsion diagrams 

 for ductile metals the elastic limit is very well defined. 



Similar diagrams to these, but taken from much larger 

 specimens, by Mr. Charnock, at the Bradford Technical 

 College, were published in Engineering some years ago. 



It was pointed out that, by means of the Richie-Gray 

 automatic apparatus, it is possible to draw a curve showing 

 the relation between stress and strain within the elastic 

 limit, in addition to the one for the complete test, the 

 extensions being magnified several hundred times. Dr. 

 Kennedy has produced diagrams similar to these by 

 making use of the apparatus already described (Fig. 106, 

 p. 222 ante), in a somewhat modified way. When a 

 diagram of this kind is to be drawn with the apparatus, 

 the spring-piece is dispensed with, and the instrument 

 fixed upon the specimen itself, by which it is supported. 

 In this case the swinging pointer does not record the load, 

 but the magnified stretch of the specimen. The indications 

 of load variation are obtained direct from the poise weight, 

 whose movement is transmitted to the sliding glass plate 

 upon which the diagram is to be drawn. There is not 

 here the same objection to taking the load from the poise as 

 in the former cases, because the loads and strains are both 

 progressive and there is no going back of the stress ; there 

 is no difficulty in keeping the weighing beam floating 

 throughout the drawing of the diagram. An example 

 of a figure produced with this device is shown on* 

 Fig. 119. The test was made upon a sample of cast iron 

 f in. diameter and 10 in. long. Three curves are shown, 

 taken one after the other from the same bar. The zero 

 line for extensions is A B. The first diagram taken is 

 marked L. When a certain point was reached the load 

 was removed, and the needle went back, not to the 

 point A, but to a higher point C, the distance A C 

 representing the permanent set after the first loading. 

 The process was now repeated, the curve starting from D 

 and almost coinciding with the first curve at the end. A 

 smaller amount of permanent set is shown in this case. A 

 third diagram was drawn, this being almost identical with 

 the second. 



It must be remembered that the extension ordinates 

 produced with this apparatus are on circular arcs, and the 

 whole curve must be replotted when it is desired to 



* Taken from Professor Kennedy's paper in Min. Proc. Inst. Civil 

 Engineers, vol. Ixxxviii. 



B2 



