128 DEPARTMENT BULLETIN 1216, U. S. PEPT. OF AGRICULTURE 
of a heat-treating furnace, the depth of impression under load, or after the 
release of load, can be measured more expeditiously than the diameter of im- 
pression. The depth of impression is usually determined from the relative 
motion of the plunger carrying the ball and the specimen. This method of 
determining Brinell hardness is not to be regarded as a standard method, and 
when requirements of speed necessitate its use, the Brinell hardness number 
should not be determined from the geometrical relation of depth of impression 
to diameter, based on the assumption that the impression has a spherical 
surface. In such tests the relation between the Brinell hardness number and 
depth of impression may be determined with a fair degree of accuracy from 
tests on specimens of material of different degrees of hardness, several impres- 
sions being made on each specimen, and the depth under load, or the depth on 
realeased of load, and the diameter of impression after release of load being 
measured for each impression. By the use of a diagram or by interpolation, 
a table giving the Brinell hardness number for a given depth of impression can 
be prepared. 
CALIBRATION OF BRINELL HARDNESS TESTING APPARATUS 
10. A load-measuring device which is in common use for Brinell hardness- 
testing machines involves the measurement of fluid pressure on a piston of 
known diameter, working in a cylinder with so perfect a fit that packing is not 
necessary. To this piston is attached the ball. Other load-measuring devices 
use combinations of weights and levers. The use of dead weights acting on a 
small piston is to be preferred to the ordinary Bourdon gauge for measuring 
the intensity of fluid pressure. 
11. The load-measuring device of a Brinell hardness-testing machine may be 
calibrated by the use of dead weights, or by the use of dead weights and 
proving levers or by an elastic calibration device or spring, in the manner de- 
scribed in the standard methods of verification of testing machines (serial 
designation (3, 13) E 4. seep. 105) of the American Society for Testing Materials. 
12. The load-measuring device of a Brinell hardness-testing machine may be 
calibrated by making a series of impressions on specimens of different degrees 
of hardness, measuring the diameters of the impressions, making a second 
series of impressions by the use of any standardized testing machine and 
standard steel ball, and, using the same measuring device as for the first series 
of impressions, measuring the diameters of this second series of impressions. 
The error of the machine under calibration can be determined by the relative 
average values of the hardness numbers determined by its use and by the use of 
the standardized testing machine. 
13. When the above companion-impression method of calibrating a Brinell 
hardness-testing machine is used it is recommended that the specimens used 
be approximately 1 by 1 by 12 inches in size, and that specimens with as wide 
a range of hardness as is feasible be used. It is recommended that impres- 
sions made by the machine under calibration be spaced about 2.5 inches apart 
along the specimen, and that the impressions made by the standardized testing 
machine be spaced alternately along the length of the specimen with the 
impressions made by the machine under calibration. 
14. If any determination of hardness, either for the machine under calibra- 
tion or for the standard machine differs by more than 2 per cent from the 
mean of the four determinations made with that machine it is recommended 
that such determination be discarded. If two determinations differ more than 
2 per cent from the mean, the whole series should be discarded. 
1". a Brinell hardness-testing machine is acceptable for use over a loading 
range within which its load-measuring device is correct within 3 per cent. 
16. The micrometer microscope or other device for measuring diameter of 
impression may be calibrated by using it to measure lengths on a standard 
steel or glass scale. The eyepiece of a micrometer microscope should be ad- 
justed so that throughout the range covered the error of reading does not 
exceed 0.01 millimeters. 
