analysis; of pull and ^^^G.\■l•:TI/.ATI()^• MKAsruK.MKNTs 95 



oi <p. lloiv tlic iiici'casc in flux dcusily in the iion parts cxtcnial lo the 

 core results in a si<>;iiificant increase in (»{/,■ us ^p incr(>ases. Jiy comparison 

 with the plot of (»{,• included in Fijj;. 1, however, this increase in (({a- is 

 minor, and does not affect th(> fact that the limit in_ii reluctance is 5/V", 

 where tp" is the satui'ation flux of the core. 



It may be noted in passinj>; that the values of 6{k lyinji; on ip' must 

 conform to (o). As this is of the same form as (3), these \-alues of ni^. 

 may be used to evaluate the component terms of (5) by the procedure 

 applied to the values of (R' in Fig;. 7. .Vs the same data are employed, 

 the \alues of (R02 , (R/,2 , and A^ thus determined agree with thos(> com- 

 puted by means of equations (7) within the accuracy of th(> computa- 

 tions. 



Case of AniialKrc SalKration 



The analysis of the low density relations is, of course, independent of 

 where saturation occurs, and involves merely the determination of the 

 e([uivalent magnetic circuit constants by the procedure described above. 

 In the high density region, however, the (luantities appearing in Fig. 6(a) 

 must be evaluated when armature saturation occurs. This recjuires 

 measurements not only of JF and ip, but also of 'Je and tpA . 



■Je is given by the Ellwood mmf gauge measurements previously 

 described. To measure <pa recjuires the use of a search coil wound on the 

 armature, located over the region of maximum density. For twin return 

 path structures, such as the relay of Fig. 2, twin search coils must be 

 used, connected to measure the total armature flux. If the variation of 

 (ft.i with (Pa is to be determined directly, additional mmf gauge measure- 

 ments must be made of the potential drop 6{a<pa through the armature. 



With (Pa determined as a function of ^e for various values of x, the 

 reluctance (Rf , or Se^<Pa may be analysed by the procedure described 

 abo\-e for the analysis of the reluctance (R or iT/V- The cur\-es of (R^ versus 

 ^E have minima (R^- at (Pa . The reciprocals of these values of (Rp are 

 plotted and analysed as in Fig. 7 to evaluate the e(iuiAalent constants 

 Ol/.p , (R02 , and Ao of Fig. 6(b). The constants (R .4 , (Pa , and ^.4 characteriz- 

 ing the relation between 6\a and (Pa are determined either from mmf 

 gauge measurements or by the method of Fig. 9. Then the constants 

 <^R/.-4 , (Ro3 , and .4:} of Fig. 6(a) can be evaluated by means of the e(|uiv- 

 alence equations (7A). 



To complete the determination of the (luantities appearing in Fig. 6(a), 

 (R/.r is e\-aluated as the ratio ^e ((P — s^i), while (*{,■ , which may eithei- 

 be constant or conform to (6), is e\aluated from the relation between i^ 

 and ;7c . 



