METHODS OF. SAMPLING AND TESTING HIGHWAY MATERIALS 111 
strength of the specimen is approached. It is recommended that a mark be 
placed at the middle of the trig loop and that a pointer be placed on the beam 
for verification tests, and that the pointer be kept as nearly on the mark as 
possible. 
21. (a) The average tensile strength (in pounds per square inch) of the odd- 
numbered specimens, multiplied by the nominal area of cross section, shall be 
considered a measure of the indicated load, and the average tensile strength 
of the even-numbered specimens, multiplied by the nominal area of cross section, 
shall be considered a measure of the applied load. The error shall be computed 
from these data. 
(b) If the tensile strength (in pounds per square inch) of any verification 
specimen varies more than 1 per cent from the mean of the group with which 
it is tested, its tensile strength shall not be used in computing the error of the 
testing machine. If more than one specimen in the odd-numbered group or one 
in the even-numbered group exceeds the allowable variation in tensile strength, 
all the specimens shall be discarded and another set prepared. 
TOLERANCES FOR TESTING MACHINES THAT MEASURE LOAD 
22. All testing machines that measure load shall conform to the following 
tolerances : 32 
(a) New machines. — The error for loads in the loading range shall not exceed 
±1.0 per cent for new machines or for those which have had any of the knife 
edges in the weighing mechanism repaired or renewed. 
( b ) Used machines. — The error for loads in the loading range shall not exceed 
±1.5 per cent. 
23. The indicated load of a testing machine shall not be corrected either by 
computation or by the use of a calibration diagram to obtain values within the 
required tolerance 
TIME INTERVAL BETWEEN VERIFICATIONS 
24. It is recommended that testing machines, when in constant use, be verified 
at intervals of six months and, when used intermittently, at intervals of two 
or three years. Testing machines shall, however, be verified immediately after 
making repairs or adjustments of the weighing mechanism, and whenever there 
is reason to doubt the accuracy of the results, without regard to the time interval 
since the last verification. 
REPORTS AND CERTIFICATES 
25. A clear and complete report shall be prepared of each verification of a 
testing machine. This report shall contain all data necessary to verify the 
accuracy of the report, and shall be filed for future reference. This report shall 
state clearly the method of verification used. 
26. A certificate giving the maker's serial number and a brief description of 
the testing machine, the makers name, the date of verification, and the loading 
range shall be signed by the person responsible for the maintenance of the test- 
ing machine, and this certificate shall be posted in plain view of a person operat- 
ing the testing machine. 
68. TENTATIVE METHODS OF TENSION TESTING OF 
METALLIC MATERIALS 53 
(A. S. T. M. tentative standard method. (5) Serial designation E 8-27 T) 
These methods deal with the form and dimensions, the machining, and the 
testing of tension test specimens of metallic materials. Such matters as methods 
52 It is believed that the tolerances given in this section may be made much smaller when 
more perfect methods of verification become generally available : and that the tolerances 
might well be materially reduced if there is available equipment for verifying testing 
machines by the use of standard weights (sec. 4) or by the use of standardized proving 
levers (sees. 5 to 8). 
53 Criticisms of these tentative methods are solicited and should be directed to H. F. 
Moore, chairman of subcommittee of committee E-l on mechanical testing. University of 
Illinois. Urbana, 111. These tentative methods are in effect a revision of the methods' for 
tension tests of the standard methods of mechanical testing of metallic materials. The 
standard methods, which were last published under the serial designation E 1-18, have 
accordingly been withdrawn. 
