SHEET NON-FERROUS METALS 



757 



may result in variations in readings which are not apparent without 

 comparisons with standard blocks. A set of directions for calibrating 

 the Rockwell tester was drawn up which called for a thorough check 

 of these various features of the machine. (See Appendix I.) 



Other questions had to be settled, such as the size of ball pene- 

 trator to be used and the amount of the load. The possibility of vary- 

 ing the load and size of ball renders the Rockwell machine capable of 

 broad application for testing purposes. Preliminary tests made to 

 determine the most suitable combination of ball and load to be used 

 indicated that the standard "B" scale of the instrument (re^-in. ball 

 and 100-kg. load) would be the most satisfactory for brass. However, 

 it was decided to try also a |-in. ball with the same load. In our 



30 



^^20 



u-K 10 



+ MACHINE G 

 o 'I H 



• 'I F 



A II E 



10 20 30 40 50 ^ 60 



ROCKWELL HARDNESS, MACHINE D — i^'bALL 



Fig. 3 — -Preliminary Comparison of Rockwell Testers. 



70 80 90 100 110 



100 KG. LOAD -RED FIGURES 



later work on nickel silver and phosphor bronze it was found that 

 the 150-kg. load gave more satisfactory results with those materials. 

 A larger ball is desirable to get readings on annealed brass and copper 

 but with harder materials sensitivity of reading is increased by 

 increasing the load and retaining the i^-in. ball. 



The question of the number of thicknesses of material to be tested 

 was also given consideration. On very thin materials there is a 

 distinct anvil effect which can be reduced or eliminated by super- 

 posing several samples on each other. However, this method of test- 

 ing soon proved to be objectionable. The Rockwell tester does not 

 distinguish between actual penetration of the material and descent 

 of the penetrator from any other cause, and since one hardness number 

 of the Rockwell machine corresponds to an 0.00008-in. movement 

 of the penetrator, great care must be taken to insure that movement 

 of the penetrator is due only to actual penetration of the material. 

 Unless the spaces under the penetrator between two or more layers 

 are completely closed by the 10-kg. minor load, low readings will be 

 obtained. With spring materials, which are frequently curved due 



49 



