482 



REPORTS ON THE STATE OF SCIENCE. — 1919. 



Tension Tests of Tubes. 



Compression Tests. 



Material Form 



Spruce Tube o.d. 1-36 inch 



i.d. -75 „ 

 Aluminium Solid 1-44 inch diam. 

 alloy 



,, )» l-o5 ,, ,, 



Mild steel Solid, annealed 



-78 inch diam. 

 Aluminium Solid 1-07 inch diam. 

 alloy 



Steel Streamline Tube 



Method of Loading 

 Through ball on steel cap 



Maximum Strain 

 Mean Strain 

 1-205 



Through ball on turning centre 



Tlirough ball on steel cap 



,, ,j ,» ,, 

 Through ball on turning centre 



Tlirough ball on steel cap 



1-058 



1-706 

 1-110 

 1-142 



1-096 



1-362 

 1-362 



The results show tliat the variation of strain, and therefore of stress, 

 across the section of a test piece is usually appreciable and is sometimes 

 of considerable magnitude. 



For plain ended tension specimens of mild steel held by wedge grips, 

 the maximum exceeds the mean strain by an average value of 16 per cent., 

 reaching 24- 2 per cent, in one case. For cast stsel, similarly gripped, the 

 corresponding figures are 128 and 16 per cent. The serrated wedges 

 appear to bite more deeply into annealed mild steel, and the figures ris3 

 to 40 8 and 62 per cent. This inequality is so great that it is rather 

 startling, but it is reduced for similar specimens to 9'9 and 16'1 per cent, 

 by the adoption of scrnved ends. There is no evidence of a corresponding 

 inaccuracy due to wedge grips after a high carbon steel is annealed. 



Eod brass and aluminium bronze gave results which are in general 

 agreement with thos^ summarised above, but in one case the bronze 

 with screwed ends showed maximum strains 36 per cent, and 39 per cent, 

 greater than the m»an. It appears probable that here the inequality 

 is due rather to lack of homogeneity of the material than to eccen- 

 tricity of loading, and the same explanation may be given to account for 

 the variations with the accurately constructed Robertson loading device 

 used in some of the experiments. 



The Effect of Eccentric Loading on the Elastic Limit and Yield 



Point. 



Eccentric loading would be of no importance if it did not affect the 

 test results. The elastic-limit, yield-point, and maximum stresses should 

 all be considered to determine the effect upon them. It may be noted 



