COMPLEX STRESS DISTRIBUTIONS IN ENGINEERING MATERIALS. 409 
TABLE VIII. 
ALTERNATING BENDING. 
y - P 
Values of = 
F 
where p, = skin stress at ultimate stage of test calculated for a distribution of 
stress as in fig. 18. 
Pe = skin stress at ultimate stage of test calculated on usual assumption 
of linearity of stress from axis throughout specimen. 
Value of “2 
| e 
Specimen | | 
for for for for for for 
KE’ i’ Ek’ 10 i’ i’ os 
. 995) — —0- — = = 
moe B= 0718 | =O 225 E 0:3 E 0-4 E 5 
| Al4 hollow 076 0-81 0-83 0-845 0-87 0-905 
| ALG. ,, 0-81 0-84 0-855 0-86 0-89 0:915 | 
| All solid | 0-67 0-735 0:75 0-785 0:83 0:87 
|) Al dee. 55 0-67 0:72 0:73 0-755 0-79 0-83 
VI. 
The Repeated Bending of Steel Wire. 
By Waurter A. SCOBLE. 
As a preliminary to the testing of complete wire ropes, single wires were taken 
from the cables and tested by repeated bending, under several tensions, over pulleys 
of different diameters. 
The wire was special acid patent steel left black. Two sizes of wire were used, 
of 0-021 and 0-036 in. diameter. The elastic limit stress was 64 and the breaking 
stress was 85 tons per sq. in. 
Under test a wire passed over a freely running pulley and was given a reciprocating 
motion. It passed from the straight on to the pulley, then it moved back into the 
straight again. Pulleys of different diameters were used. Direct tension could be 
applied to the wire under test, and experiments were made at several tensions on 
each pulley. 
The wire was stressed by bending it to the radius of the pulley, and it was 
anticipated that an additional direct tension would reduce the number of bends 
necessary to produce fracture. The endurance of stranded wires is reduced on a 
given pulley if the tension is increased. 
The results obtained from the 0-021 in. were confirmed by the tests of the 0-036 in. 
diameter wire, therefore attention will be directed particularly to the former. 
The usual formula used to obtain the outside fibre stress in the bent wire is, f = Ed/D, 
and if ‘ f’ be taken as the yield stress of the wire in tension, 64 tons per sq. in., and 
* E’ as 13,400 tons per sq. in., it appears that the wire would be stressed to its yield 
point when bent on a pulley of 4-4 in. diameter without an added longitudinal 
tension. Under simple tension this wire yields at 50 and fractures at 66 lb. approxi- 
mately. 
The experimental results may be divided into four sections according to the pulley 
diameter. On a pulley which was much too small for the wire the number of bends 
to fracture was low but approximately constant up to a particular tension, above which 
the number of bends was negligible. This is illustrated by : 
