﻿522 



Mr. S. Lees on a Simple Model 



exceed numerically a certain amount, and when this does 

 occur, the area is proportional, for small-sized loops, to the 

 excess of F x over the critical amount. Also, immediately 

 after a change in sign of dFjdt, where F represents stress 

 and t the time, the stress-strain curve is always straight and 

 parallel to the straight line which represents elastic change 

 without hysteresis. 



On the other hand, the author's theory fails to explain two 

 noticeable points indicated in Messrs. Smith & Wedgwood's 

 paper : — 



(1) When a point corresponding to K in fig. 4 has been 

 reached, the stress-strain curve is actually found to become 



Fig. 5. 



curved, as shown in fig. 5, instead of following a straight 

 line like KL of fig. 4. 



(2) For large areas of loop, it is actually found that the 

 proportionality between area of loop and excess of F x over 

 the critical amount above referred to breaks down. 



Other divergences between experiment and the theory at 

 present outlined will be indicated below (§ 9). 



§ 7. Stead y Hysteresis Loop for Unsymmetrical Stress Limits. 



We shall now discuss the nature of the steady hysteresis 

 loop when the limits of stress are not equal and opposite. 

 The appropriate stress-strain (in our case, F— $) diagram is 

 indicated in figs. 6 and 6 a. From the considerations out- 

 lined in § 4, it will readily be seen that just previous to F 

 reaching its lowest value (as at J), the F — x relation will be 

 (on the theory outlined) d¥/dx=X. The corresponding line 



