ON STRESS DISTRIBUTION IN ENGINEERING MATERIAI.S. 485 



The points referred to are, briefly stated , as follows : — 



(1) The effect of the variation of tensile stress as distinguished from 



a variation of stress from tension to compression. 



(2) The variation of the maximum limit when combined with a fixed 



lower limit of range of stress. 



(3) The limiting range of stress and its relation to the maximum stress, 



its effect on the elastic properties of the material, and the 

 connection between complete recovery and perfect elasticity. 



(4) Any indications given by the crystalline structure as to whether 



failure is likely to occur. 



The testing machine was specially designed for the setting up of tensile 

 stresses only by the late Professor T. C. Fidler, M.I.C.E., and was con- 

 structed in the Engineering Department of University College, Dundee. 

 The load is applied by means of hydraulic pressure acting on the underside 

 of a piston to which is attached through a spherical bearing joint the upper 

 end of the test piece, the lower end being similarly connected to a fixed 

 cross-head. The number of alternations of pressure is capable of being 

 varied from 10 to 18 per minute. 



A special feature of the machine is its ability to take test pieces of the 

 same dimensions as those ordinarily used in a Wicksted Machine with a 

 length sufficiently great to allow of the use of a Swing's Extensometer. 



The material used in the first test pieces was Siemens Martin Steel, 

 and under static load showed a yield stress of 15 tons per sq. in. and a 

 maximum stress of 26-4 tons on the original section with an ultimate 

 elongation of 32-8 per cent, on an 8-in. length. The later specimens, 

 in order to fit in with the scheme of research framed by the Engineering 

 Committee of the British Association on Complex Stresses, were of specially 

 supplied Dead Mild Steel (-12 per cent. Carbon), yield stress 15'5 tons per 

 sq. in., maximum stress 23-4 tons, and elongation 33- 7 per cent. 



In the series of tests, in which alternations of tensile stress only were 

 given, it was found that : — 



(1) After ordinary yield is exceeded and permanent elongation has 



taken place, the limits of the proportionality of stress and 

 strain are raised. The extent of this elongation for any 

 maximum stress dejDends on whether the stress has been 

 apj)lied in one stage or with a sufficient number of repetitions 

 at each of several stages. 



(2) The range of stress over which this proportionality is maintained 



is limited ; once established it is independent of the number 

 of cyclical repetitions to which it is subjected. 



(3) The range of stress may be again varied by further elongation, 



but to this there is a limit. 



(4) Failure under repetitions of tensile stress occurs by the move- 



ment of the crystals relatively to one another, consequent 

 on the existing range of stress being exceeded, or during the 

 transition period when one range is being changed to another. 



(5) Failure frequently originates by the range of stress, or even 



the maximum static stress of the material, being exceeded 

 in parts of a section, and is due to mechanical flaws in the 

 material or to non-axial loading. 

 1919. N N 



