Ductile Materials under Combined Stress. 75 



ticular kind of test, it is unaffected, or even increased, for tests 

 involving the application of stresses of a different type ; 

 examples of this may be seen in figs. 16 & 18, in which a 

 tension and a torsion test respectively upon the material are 

 compared, the material in the interval of the two tests having 

 been subjected to yield-point stresses of a different type. 



Not very many years ago the existence of the yield-point 

 was not recognized, the stress-strain curve sketched as normal 

 shoving a gradual bend over (somewhat similar to the copper 

 tension tests shown in figs. 21-24). This would be the case 

 with imperfect material, the portions of the material in the 

 neighbourhood of flaws and variations reaching the yield - 

 point successively, and so masking the phenomena. There is 

 no doubt that the materials of construction as produced to-day 

 are much more uniform and free from defects than those of a 

 quarter of a century ago, and thus can display phenomena 

 perhaps hidden in the past by their imperfections. 



It should be noted that the same effect (elastic-limit effect) 

 on the stress-strain curve is caused by the non-coincidence of 

 the axis of the specimen with the load-line. 



8. The Yield-point, in preference to the Elastic limit, selected 

 as the criterion of strength. — The elastic-limit effect may 

 then be merely due to local yielding, and taking this view of 

 it I have regarded the yield-point as the true criterion of the 

 strength of the material, and sometimes have assumed chat 

 Hookers law held up to it. 



In all those of the experiments of this series which were 

 made upon steel, a- yield-point, sometimes much obscured by 

 elastic-limit phenomena, was sought for and determined ; and 

 evidence will be given to show that, although there is on the 

 whole a slight tendency for an annealed specimen, when 

 frequently stressed, to have its yield-point rise, yet, if the 

 material be not much strained at the yield-points, this rise is 

 not very large, especially if the type of stress be varied. In 

 the case of copper, the curves obtained were such as to make 

 it difficult to determine the yield-point, if indeed one exists ; 

 several of the curves, however, have been plotted out so that 

 the range of choice may be appreciated. 



9. Torsion, a case of Combined Stress. — The case of torsion 

 presents a case of combined stresses, two principal stresses 

 being equal in amount but opposite in sign, while the third 

 is zero. Of such tests a large number have been made on 

 round bars, but the essentials of the phenomena involved are 

 masked by the variation of strain from the axis outwards, 

 which causes the outer layers to reach the yield-point first, 

 and so produce effects corresponding to the explanation above 



