SECTIONAL TRANSACTIONS.—G. AL5 
ever since, and a great many interesting facts have been discovered, though 
the fundamental problem—what the real cause and nature of fatigue failure 
may be—has not yet been solved. 
The work of the panel includes investigation on the following subjects : 
The existence of a finite fatigue limit; the mathematical explanation of how 
a fatigue crack extends; the nature of elastic hysteresis; phase changes in the 
metal ; the relations between the directions of the principal stresses, the crystal 
axes and the plane of the fatigue failure; the effects of temperature on the 
fatigue limit; the action of fatigue at points of stress concentration; fatigue in 
single crystals; fatigue in amorphous material; the strengthening of metal by 
fatigue, and many others. 
10. Prof. F. C. Lea.—The Effect of High Temperature on the Range 
of Repetition Stress for Steels. 
The apparatus used in connection with the Haigh repetition stress machine 
is briefly described. The endurance range for equal compressive and tensile 
stresses and for unequal ranges of stress at temperatures varying from 0° C. 
to 800° C. are given. The phenomenon of creep during statical tests and also 
during repetition tests is discussed. The fatigue range, for 10 million repeti- 
tions, at temperatures as high as 500° C., for equal tensile and compressive 
stresses, is shown to be higher than at ordinary temperatures; the strength of 
these temperatures under a statical applied stress is shown to be very low. 
For unequal stresses it is shown that creep may continue for very many million 
repetitions, and afterwards the fracture is a ‘ fatigue’ fracture. 
11. Messrs. H. F. Gouau and H. J. Tapseun. 
Fatigue Tests. 
Some Comparative 
12. Mr. C. E. Srromever.—Torsion Fatigue Hysteresis. 
Tuesday, August 12. 
13. Prof. T. M. Jasper.—Measurement of Quenching Stresses in 
Steel. 
14, Prof. H. P. Pumpror.—The Dimensional Problem and Signifi- 
cance of the Notched Bar Test. 
The paper deals with experiments made with the object of finding the type 
of equation connecting the energy absorbed in the notched bar impact. test 
with the dimensions of the test piece. ‘The test piece is tested in the single- 
blow pendulum machine and is notched on one side; it is bent by the blow 
away from the notch, which is set at the level of the top of the vice. The 
notches used were of the form adopted by the Royal Air Force, and subse- 
quently standardised by the British Engineering Standards Association. The 
test pieces were all cut from the same bar of heat-treated nickel-chrome steel. 
The breadth of the test piece 6, the thickness behind the notch t, and the 
distance 7 of the striking edge. of the hammer above the notch, were varied 
one at a time, giving 19 variations of these three dimensions, and each variation 
was represented by a number of tests. ‘lhe equation derived is of the form: 
Energy absorbed =c.b.t. ie +m) . (l4+n), where c, m and n are constants. This 
equation is shown to be of rational form provided that ¢ has the dimensions 
of a stress intensity, m is a ratio and n is a length. The significance of the 
test is discussed. 
15. Report of Committee on Complex Stress Distributions im 
Engineering Materials, 
