6 THE BELL SYSTEM TECHNICAL JOURNAL, JANUARY 1954 



to the force. The crosshead motion is transmitted electrically to a servo 

 system which drives the paper up or down in proportion to the displace- 

 ment of the mechanism involved. Thus, during the motion of the cross- 

 head, the pen moves proportionately to force, the paper moves propor- 

 tionately to distance, and a force-deflection curve is drawn. As a result, 

 the tensile characteristics of mechanisms can be recorded very easily. 

 For example, force variations of relay springs or of relay magnets may be 

 measured, or a combination of the two, as desired. 



Some typical measurements obtained on this machine are given in 

 Fig. 2 showing on one chart the manner in which electrical contacts 

 change their force against the magnet and the manner in which the 

 magnet pull varies across the gap. Such charts as these, when completely 

 analyzed, enable the designer to choose the proper magnet for a particu- 

 lar spring requirement. Among the useful features of this machine are 

 the accuracy of recording which readily provides one or two per cent 

 accuracy, and the ease of tracing and retracing a particular measure- 

 ment. Also I'eadily obtained are exact information on the mechanical 

 hysteresis losses within the mechanism. For example, in curve 1 of Fig. 

 2, the load characteristic of the contact springs is seen to have two 

 values. The upper one represents the force on the closure stroke while 

 the lower one represents the force on separation. The area between these 

 two curves is the hysteresis loss; friction at any point is readily estimated 

 as one-half the difference betw^een upper and lower force readings. Special 

 information, such as the exact location of first closure or first separation 

 of contacts can also be included, as the machine is provided with a circuit 

 to insert a "pip" on the record w^hen desired. 



This instrument has marked advantages because of its accuracy, speed, 

 convenience and wide range of uses. With almost equal ease, mechanisms 

 whose full force variation is 2 grams, and those ranging up to half a ton 

 can be measured. Once a suitable jig has been prepared to properly 

 mount the parts, the Instron machine will furnish a complete set of data 

 in less than an hour, whereas by previous methods it would have re- 

 quired a matter of days. 



The time required for each indi\'idual cur\'e in the figure is the time 

 allowed for the crosshead to move through the relay stroke. This usually 

 is about 30 to 60 seconds, which adequately simulates the static charac- 

 teristics of the relay force system; i.e., those force properties that would 

 be measured at a particular point if the armature were held there. These 

 static characteristics are commonly used in relay design to establish per- 

 formance criteria for ordinary relays. However, in detailed studies of 

 relay dynamics, these static characteristics must be supplemented by 



