DYNAMIC MEASUREMENTS ON ELECTROMAGNETIC DEVICES 1461 



The true decay cun-e has an initial steep slope followed by continuous 

 curvature. A better approximation is to stipulate an initial jump dis- 

 continuity. A good fit to the flux range in which release of «>U'.t inmagnets 

 occurs, shown by the lower straight line, is 



I = .601 e"'^'" t ^ 0. (20) 



This discontinuity of flux in the first approximation is just the reverse 

 of the continuity of flux concept usually used. Of course, no actual dis- 

 continuity occurs, as shown by the true decay curve. The failure of any 

 single exponential equation to represent the true curve merely makes 

 clear the fact that the behavior of the core is not that of a single coupled 

 turn, but rather is that of an infinite line. Bozorth^ gives the intercept 

 as 0.691 for the linear case, which does not represent the continuously 

 curving decay which actually occurs. 



The chosen intercept of the first approximation curve at / = is ad- 

 mittedly somewhat arbitrary. It was arrived at in a broader study in- 

 cluding flux rise curves. Accepting this ec^uation, a convenient determina- 

 tion of te can be made similar to that for exponentials. If t is set eciual to 

 te, then 



= 0.221. (21) 



Thus, after measuring a dynamic flux decay curve, the time can be 

 determined for which the above ratio obtains. This directly is te. From 

 linear circuit theory and Lenz's Law, the inductance for one turn is: 



i. = r/. (22) 



whence Ge can be determined. 



Now because of magnetic saturation and the shape of the hysteresis 

 loop, the values will depend upon the particular final ampere turns (NX) 

 used in the experiment. For uniqueness and uniformity in rating electro- 

 magnets for comparison purposes, the particular set chosen is that for 

 which Li is a maximum. For comprehensive operating studies, measure- 

 ments of course have to be made under the actual conditions of interest. 

 Except for rating purposes, t^ is not ordinarily split up into components. 



Thus, while the rated value of Ge for a particular electromagnet has 

 the dimensions of conductance, it includes other factors as well. Some are: 

 (a) Core material conductivity, (b) Magnetic non-linearity, (c) Shape of 

 the hysteresis loop, (d) Non-uniform flux distribution, (e) Eflfect of pole 



