756 BELL SYSTEM TECHNICAL JOURNAL 



is much smaller than that of the Rockwell on the same samples. In 

 other words, the sensitivity of the scleroscope is considerably less 

 than that of the Rockwell. 



An investigation was made of the type "D" recording scleroscope 

 equipped with a universal hammer to obtain a comparison with the 

 type "C" instruments. The conclusion reached was that there was 

 little difference between the two instruments as regards precision. 

 It was not considered worth while, therefore, to make elaborate com- 

 parison tests. Hence, this type of instrument was not used in the 

 round-robin tests. 



In view of the difficulties encountered in using the scleroscope, 

 which have also been brought out by previous investigators, no 

 further effort was made to adapt it for use as a specification instrument. 



Rockwell Hardness Tests: 



While the Rockwell hardness tester has been used quite extensively 

 and with considerable success in testing steels it has been used com- 

 paratively little for testing non-ferrous metals. Our experience 

 previous to this investigation indicated that it might prove satisfactory 

 as a specification instrument for non-ferrous metals. Considerable 

 work had already been done by various laboratories to determine 

 the limitations of this machine. Furthermore, the routine testing 

 with the Rockwell of all incoming shipments of material had been 

 instituted for the purpose of accumulating specification data. While 

 each individual machine seemed to give satisfactory results to the 

 user there was little information available as to what agreement 

 could be obtained with machines in other laboratories. Figure 3 

 gives results of tests on identical samples with five different machines 

 in the laboratories of one of the participating companies before any 

 attempt was made to eliminate mechanical irregularities in the 

 machines. While there was considerable difference between machines, 

 there was a probability that these differences could be reduced by care- 

 fully going over the machines and establishing technique of test. 



Before making any further comparative tests, careful study was 

 made of each machine and various mechanical irregularities were 

 eliminated. In order to get close comparative results the ball pene- 

 trators must be in good condition, the load must be applied without 

 impact and at approximately the same rate in different machines, 

 the value of the load must be the same, and the method of supporting 

 the specimen across the anvil is important. Due to the lever system 

 of applying load and measuring the amount of penetration of the 

 material under this load, a sHght amount of friction in the bearings 



