480 THE BELL SYSTEM TECHNICAL JOURNAL, MAY 1952 



shown by Fig. 9 of the next section and are discussed there. The sensitiv- 

 ity of this type of unit is higher than that for the normal force measuring 

 unit. As shown in the appendix, the voltage generated is independent of 

 the area of application and is about 9.7 volts for a unit the same size as 

 discussed above which gave 2.7 volts for 100 grams applied at a point. 

 By placing weights on the upper surfaces both types of units can be 

 used as accelerameters. They are cemented to the surface whose ac- 

 celeration is to be measured and the force applied is equal to half the 

 mass of the ceramic plus the added mass times the acceleration. By put- 

 ting weights on the shear pickup ceramic types, tangential accelerations 

 can be measured in the direction of the poling. By using three such 

 accelerameters, the normal and two tangential components of accelera- 

 tion of any surface can be measured. 



III. METHODS FOR INVESTIGATING CAUSES OF WEAR 



Wear in various parts of a relay is the limiting factor when a very 

 large number of relay operations are desired. This wear opens up the 

 spacing between contacts and causes the relay to lose its adjustment 

 over a course of time. 



A. Force Measurements and Wear Caused by Normal Forces 



Since the forces operating on a material can be divided into normal 

 and tangential forces, it appears desirable to separately determine the 

 effects of each. Normal forces were produced by using the barium ti- 

 tanate, metal horn detail of Fig. 3. With a steel ball on the end of the 

 metal horn, and a barium titanate specimen glued to the pivoted arm, 

 the peak forces in grams are plotted against the volts used to drive the 

 titanate unit for various static forces in Fig. 8. The pattern of voltage 

 is approximately a rectified sine wave, since the ball is out of contact 

 with the measuring titanate a part of each cycle. To observe the wear 

 caused by normal forces a piece of material to be studied was glued to 

 the pivoted arm on top of the barium titanate and the force was adjusted 

 to the required value. For forces in the order of those measured in relays 

 no wear at all was observed over a period of 18 hours which corresponds 

 to a billion impacts, since the number per second is 18,000. For larger 

 impulsive forces, it was found that the result of 60-million impacts 

 against an insulator such as a phenolic was to produce a pit only a few 

 tenths of a mil inch deep by a plastic flow. Since no wear of the type 

 involved in relays was observed it was concluded that practically all of 

 the wear was produced by tangential forces. 



