COMPLEX STRESS DISTRIBUTIONS IN ENGINEERING MATERIALS, 177 
probably simpler to deal with than non-ferrous metals, or even cast iron, 
to a first approximation; but Bauschinger found that the strength of 
stone varies considerably with the proportions of the specimens, and 
that stone has no elastic limit, taking sets with small loads. Hard and 
dense stones are better in these respects, and all are better at higher loads. 
The difference in the strength of some rocks in different directions is very 
great. Nagaoka® also found rocks to be very imperfectly elastic, but 
Adams and Coker® consider their elasticity in compression to be better 
than that of cast iron, especially after they are loaded several times to 
attain a state of ease. 
The errors which are likely to be introduced in tests of rocks in com- 
pression are now well known, and the best-conducted tests leave some 
uncertainty regarding the true compressive strength. 
It is impossible to deal fully here with the behaviour of the crystals in a 
material under stress. The researches of Ewing and Rosenhain are well 
known, and those of Beilby deserve notice. The discussion on the papers 
of Mason and Smith ®* included a reference to this matter by Gulliver, and a 
most suggestive contribution from M. Osmond. 
Papers which deal with experiments made on rocks usually refer to the 
behaviour of the separate crystals. 
13. Alternating Combined Stresses. 
The only experiments with which we are acquainted which have been 
intended directly to investigate alternating combined stresses are those of 
Turner.*® The plan of the research was not all that could be desired, but 
was probably the best that could have been done with the available facili- 
ties. Specimens were tested under alternating bending and torsion, but 
not combined. The torsion was taken as an example of combined stresses. 
The chief results are shown in the table. 
Material Tube Steel) Mild Steel | Tool Steel Nickel Steel — 
Ib./sq. in. | 1b./sq.in. | 1b./sq.in. | 1b./sq. in. 
Elastic Limit . : 31,000 42,300 67,000 creat | Tension by 
Endurance a 29,000 40,000 50,000 59,000 bending 
Elastic Limit . ; 17,000 | 24,000 38,400 40,800 } Shsar 
Endurance A é 16,000 22,000 38,000 35,000 
Per cent. Elong. (8) 24 | 29 9 14 | At fracture 
The tube and mild steel specimens conformed to the shear stress law 
under alternating stresses. The tool steel was particularly weak in 
alternating tension, and with nickel steel the drop in strength was greater 
in tension than under torsion. The percentage elongations indicate that 
the more ductile steels obey the shear stress law under repeated loadings, 
and the behaviour of the more brittle samples approaches more closely to 
that required by the maximum strain and maximum stress theories. 
Wohler tested steels under repeated tension and compression, bending, 
and torsion. It is difficult to compare his results for our purpose, since there 
were considerable differences in the material included under the same 
title—the elongation at fracture for Krupp’s cast steel for axles varied 
from 11-7 to 23-7 per cent. The range of stresses which he selected from 
all his tests probably refer to an average sample, and these for cast steel 
1913. N 
