760 



BELL SYSTEM TECHNICAL JOURNAL 



The effectiveness with which standard test blocks may be used in 

 bringing Rockwell testers into agreement is shown by the following 

 calibration : 



Before overhauling and checking the entire scale, 

 average difference in hardness readings between 

 three testers 



After overhauling and checking the entire scale, 

 average difference in hardness readings between 

 three testers 



Approximate Hardness, "B" Scale, 



^i^-in. Ball, 100-kg. Load 



(Red Figures) 



25 



6.0 

 0.9 



60 



3.7 

 1.2 



80 



3.4 

 0.9 



Calibration of Test Blocks. — It was first thought that a Rockwell 

 hardness machine would hold its adjustment over a long period if 

 carefully maintained and could serve as a standard machine to be 

 used for the calibration of test blocks. More complete experience 

 has shown, however, that wear of the operating parts and slight 

 variations in friction caused the "standard machine" to vary slightly 

 in its readings. Naturally, the calibration of standard test blocks 

 under these conditions is unsatisfactory because it might result in the 

 establishment of several sets of standard blocks, leading to confusion 

 in the application of the Rockwell hardness test limits to commercial 

 material. 



A new method of calibrating the Rockwell hardness tester was 

 devised. This method consisted in preparing a set of test blocks cover- 

 ing the range of hardness of the high-brass and clock-brass rolling series. 

 These blocks were checked by the various cooperating laboratories 

 as well as the manufacturers of the Rockwell hardness tester, namely, 

 the Wilson-Maeulen Co. These blocks were considered the basic 

 standards for the Rockwell tests. Sub-standard blocks were calibrated 

 in comparison with these standard blocks and placed in use by the 

 various cooperating laboratories. The high-brass rolling series and the 

 clock-brass rolling series were then retested for Rockwell hardness 

 using the new calibration. It was found that the hardness readings 

 differed one to two points from the values previously determined for 

 several tempers. At the same time experience with commercial ship- 

 ments of material showed that a shift was necessary in the Rockwell 

 hardness-tensile strength relationship. This verified our conclusions 

 with regard to the adoption of the new method of calibration. 



After the establishment of the standard brass test blocks it was 



