﻿Tests of Metals in Reverse Torsion. 427 



order to test the materials in the reverse direction, it was 

 necessary to devise some special attachment for the machine. 

 This finally took the form of a lever attached to the suspended 

 head, as shown in fig. 2. The lever carried a bucket at the 

 free end, to provide for the insertion of the weights. By 

 reversing the jaws of the machine and also the motion, it 

 was possible to give the specimen the desired reverse twist. 



After the specimen was properly inserted and the lever 

 and bucket in place, the beam of the testing machine was 

 balanced, eliminating from consideration the weight of the 

 parts' attached. The distance from the centre of the test 

 piece to the point of attachment of the bucket being known 

 (GO inches), made it possible to add weights to produce tiny 

 desired number of inch-pounds of moment. The loads 

 thus added caused the scale-beam to rise. By running the 

 machine in the reverse direction, the load was gradually taken 

 up by the test piece, the balancing of the scale-beam 

 indicating when all the load had been so taken up. Deforma- 

 tions were measured by means of an Olsen troptometer, as 

 seen in fig. 2, and the routine testing, in the reverse direction, 

 consisted of adding the desired weights, running the machine 

 backward until the specimen took up the load, and then 

 measuring the corresponding deformation. To test the 

 accuracy of the apparatus, series of ordinary torsion tests 

 were made, first with the arm attachment and then without ; 

 i. e,, specimens were tested in simple torsion using the arm 

 attachment, and then others were tested in torsion in the 

 same direction without using the arm attachment (in each 

 case the specimens had not previously been tested). The 

 results were precisely the same, showing the accuracy of the 

 arm attachment. 



Outline of Work. — As before stated, the tests were made to 

 determine the effect of overstrain, of the outer fibres in torsion, 

 on the elastic strength of these fibres in the opposite direction, 

 i. e., against stress of the opposite sign, or, it might be 

 stated, the effect of overstrain in tension on the elastic 

 properties in compression. To accomplish this the following- 

 plan of work was carried out : — 



(a) Bars tested in torsion to some stress below the 



elastic limit (proportional limit of bar) and the direction 



of torsion reversed immediately. 

 (6) Bars tested in torsion to the elastic limit and the 



direction of torsion reversed, (1) immediately, (2) after 



a certain time, 

 (c) Bars tested in torsion to the yield-point (of the 



bar) and then the direction of torsion reversed, (1) 



immediately, (2) after a certain time. 



