ON STRESS DISTRIBUTIONS IN ENGINEERING MATERIALS. 163 



elastic limit and yield point appeared to coincide at a stress of 

 10 tons per square inch. Judged by the first yield the steel is weaker 

 under a bending load (9*18 tons per square inch), but the point is 

 clearly defined on a carefully plotted load-deflection diagram. There 

 was a considerable creep when the skin stress would be 12 tons per 

 square inch, assuming Hooke's Law to hold. The final stress was 

 evidently between 14-2 (Hooke's Law) and 8-37 tons per square inch 

 (uniform distribution), and it appears that 9-2 tons per square inch is 

 very approximately correct for the yield point stress by bending. 

 The torsion results deserve special notice. The first yield apparently 

 took place at 3'27 tons per square inch skin shear stress, and the test 

 piece gave way and was twisting slowly but continuously when the 

 skin stress was between 6'54 (Hooke's Law) and 5'31 tons per square 

 inch (uniform distribution). The stress difference at first yield appears 

 to be much less in torsion than under bending or tension, but the 

 flow stress difference for torsion is greater than for tension, and probably 

 also greater than for bending. It appeared probable that the point on 

 the torque-twist curve where the straight portion ends was not the 

 point at which yield occurs, possibly because strain was not propor- 

 tional to stress even when the material was elastic, but such an 

 assumption is confused by the fact that strain was never proportional 

 to stress in the second torsion test. In the latter case it is im- 

 possible to locate a yield point, but the final flow shear stress was 

 between 5"71 (Hooke's Law) and 4"65 tons per square inch (uniform 

 distribution), a result which gives a stress difference in good agreement 

 with that from the tension test. 



Appendix B. 



Report on Alternating Stress Tests of a Sample of Mild Steel received 

 from the British Association Stress Committee. 



By Dr. B. P. Haigh. 

 The material supplied was in the form of -ft-inch rolled bar, and 

 was described as ' dead-mild ' : its tensile strength was approximately 

 26 tons per square inch, with an elongation of about 21 per cent, on an 

 8-inch specimen, f inch in diameter. The results of two tensile tests 

 are given below, column A giving the figures obtained in a preliminary 

 test made by Mr. Cook, and column B those in a very slow test, lasting 

 over 30 minutes, made at Greenwich: — 



Diameter of specimen (turned) 



Yield stress, tons'sq. in. ... 



Maximum stress, ditto .... 



Elongation, per cent 



Reduction of area at fracture, per cent. 



A 



0-385 

 23-95 

 26-4 



22-2 on 6 in. 

 69-8 



B 



0-374 

 21-0 

 25-2 



20-5 on 8 in. 

 71-4 



A Brinell test, in which a standard 10 mm. ball was pressed upon a 

 longitudinal section of the bar, with the standard load of 3,000 kg. gave 

 an impression having a diameter of 5'78 mm., indicating a ' hardness 

 number ' of 104 kg./sq. mm., equivalent to 66 tons per square inch. The 



M 2 



