METHODS OF SAMPLING AND TESTIXG HIGHWAY MATERIALS 115 
proportional to those shown in Figure 51 he used and that the specimen be 
made as large as feasible. In any such small-size specimen it is especially 
important that the gauge length for measuring elongation be four times the 
diameter of the specimen. 
9. In all tension tests of metals the actual dimensions of cross section of the 
test specimens shall be measured with a micrometer reading to one-two hun- 
dredths of the dimension measured, and the stresses shall be computed on the 
basis of the measured cross section : it should never be assumed that the dimen- 
sions of the measured cross section are identical with the nominal dimensions. 
Note. — It will be nored that the tolerance for diameter of specimens shown in 
Figure 51 permits the use of specimens 0.505 inches in diameter. 
/ Spherical 
Bearmg 
^Cylindrical 
Seat 
Serrated Faces 
on Grips 
Cylindrical 
Seat 
<— Specimen 
Fig. 54. — Self-adjusting grips for use with sheet metal 
10. The shape of the ends of the specimen shown in Figure 51 are not speci- 
fied above. For specimens of brittle metal it is desirable that the diameter of 
the specimen be changed gradually along its length from the minimum section 
at the gauge length to the diameter at the ends. Figure 52 shows two forms 
of end of specimen which have given satisfactory results in tests of brittle 
metals. 
11. Specimens shall be finished so that the surfaces are smooth and free from 
nicks and tool marks. All ragged edges shall be smoothed. 
