104 THE BELL SYSTEM TECHXIC.VL JOl'RX.VL, J.VXUARY 1954 



Table III — Evaluation of Magnetic Circuit Constants 



resulting plot is linear, in agreement with (20), showing the observed 

 relation between F'/F and 5" to conform fully to (15). 



The expressions for the slope and intercept given by (20) are indicated 

 in Fig. 14. From the values of the slope and intercept given by the plot, 

 Cl may be evaluated, giving (Rl , as (Ro is known. With (Rl thus evalu- 

 ated, (f" may be determined from the value of the slope. The values of 

 (Rl and cp" thus determined from results of Fig. 14 are listed in Table 

 III, together with the vsdne of (p' given by the ratio i^'/V" determined 

 from the common intercept in Fig. 13. Included in this table under 

 "Pull Results" are the \'alues of A, (Ro , and (Rl determined from the low 

 density data and prex'iously tabulated in Table II. \'alues of the same 

 quantities as determined from the magnetization measurements are in- 

 cluded for comparison. 



For both pull and magnetization analysis, empirical relations have 

 been assumed in (6) and (15) for the variation of (R with^, agreeing with 

 each other and the known conditions only at their upper and lower limits. 

 In view of this, the agreement in the values of ip' and tp" between the 

 pull and magnetization results is good, and the deviation in the value of 

 (Rl found in Fig. 14 from the other values of this ciuantit.y is not surpris- 

 ing. 



High Density Pull for Armature Saturation 



The low density pull for armature saturation may be analysed as in 

 the case of core saturation. The pull data may be plotted against tF on 

 logarithmic paper as in Fig. 11, and the tangents of slope two giving F' 

 drawn, as in this figure. The values of \f/y/F' may he plotted as in Fig. 

 12 to determine A and (Ro , and A-i and (R02 also evaluated if J^ has been 

 determined from magnetomotive force gauge measurements. As before, 

 the remaining magnetic circuit constants of Figs. 4 and 5 may be de- 



