MAGNET STEELS AND PERMANENT MAGNETS 395 



Criteria of Magnetic Quality of Permanent Magnet Steels 



A number of quantities determinable by magnetic measurement 

 have been proposed at various times by different investigators for use 

 as criteria of magnetic quality of magnet steels. Among these quanti- 

 ties may be listed (BII)max''^\ the maximum product of the coordinates 

 of the demagnetization curve; and the quantities BrIIc^^\ Br/Bmax, 

 BrIIc/Bmax aud Br/IIc'-^^. High values of all of the proposed criteria 

 except the last were considered desirable. 



The form BrjUc has decided disadvantages, since a low value of the 

 criterion is obtained when Br is small as well as when lie is large, and 

 low values of Br are distinctly undesirable in view of the fact that 

 Brem is dircctly proportional to Br. The ratio BrjBmax was found to be 

 practically constant for some types of steel, regardless of the value of 

 He, and thus was of no value as a criterion. The quantity Br Hc/Bmax 

 would accordingly amount to a constant times He and would give no 

 actual weight to the value of Br. These three criteria would therefore 

 fail either to indicate a choice between different types of magnet steel 

 or to aid in the selection of an optimum heat treatment for a given 

 kind of steel. 



This leaves the quantities {BII)max and Br He as the only ones not 

 obviously faulty. The first was proposed by S. Evershed in a note- 

 worthy paper in the Journal of the Institution of Electrical Engineers 

 for September 1920, in which he showed that the quantity {BH)max is a 

 measure of the maximum amount of external magnetic energy which 

 can be supported per unit volume of a given magnet steel, and that 

 this in turn defines the term "magnetic quality" as applied to magnet 

 steel. 



Evershed's derivation of his criterion of magnetic quality, given in 

 the article noted above, is complete and convincing, but it was thought 

 desirable to obtain experimental verification. The data from the 

 bars of Fig. 4 were used to this end. In Fig. 10 are shown a number of 

 curves each corresponding to a different type of magnet steel. The 

 demagnetization curves for the bars are shown plotted in the usual 

 way, and to the right of these are plotted the curves of {BH) vs. B. 

 To the right of these are plotted further the experimentally determined 

 curves of Brem vs. length for the bars in question. 



Fig. 10 shows that if tangents to the curves of Brem vs. length are 

 drawn through the origin the point of tangency in each case indicates 

 the value of Brem and length at which the ratio of Brem per unit length or 

 volume of steel is a maximum, and it is clearly demonstrated that in 

 each case these values of Brem coincide closely with the values of B for 

 which the product {BH) is a maximum. The writer feels that this 



