SLOW KKLEASK KKLAY DKSIGX 



199 



Similarly, the initial ii'on i<'luctaiic(> corrcspoiKls to tlH> slopo of the 

 (Icniagnetization cuinc at //,■, as <2;i\(Mi by the initial pcrmoahility m"- 

 Thus the total initial i-cluctancc (U" is ^\vv\\ by the o(iiiation: 



(ft 



(ft.; + 



1 



a 



(12) 



\\ hciv X^r a denotes the sum of terms corresponclinj"; to the component 

 parts of the iron path, each term giving the length of the part divided by 

 its cross sectional area. As m" is of the order of 20,000 for soft magnetic 

 materials, the iron reluctance term is small (•omi)ar(Ml with (Hk , which is 

 typically of the order of 0.010 cm~ or less. 



The value of <^" may be estimated as the saturation fiux for the core, 

 or B"a, where a is the cross sectional area of the core. B" is the satiu'a- 

 tion density, which may be taken as the value of B^f listed in a companion 

 article.^ Estimates of ^" thus obtained are lower than those which best 

 lit the decreasing magnetization measurements. No convenient means 

 for correcting this disparity has been found, other than an arbitrary 

 increase by a factor of 20 per cent, based upon experience. In a particular 

 relay design, however, the oV)served values of (p" vary directly with the 

 core cross section. 



PULL RELATIONS 



As sho\A'n by equation (6), the release time varies inversely as the incre- 

 mental reluctance (fti . This is proportional to 6{", in which the dominant 



- 0.020 

 5 



+ 13- 



0.010 



(R' 



0.005 



FLUX, 50, IN MAXWELLS 

 3000 4000 6000 



10,000 



- 15,000 



J + 'Jr 



IN ABAMPERE TURNS 



ATT 



I"ig. 8 — Evjilualiuii uf magnetic constants from measurements of decreasing 

 magnetization. 



