40 BELL SYSTEM TECHNICAL JOURNAL 



evidence of manufacturing and other defects were chosen. Each blank 

 chosen was cut into two 10' lengths, one of which was made into a JW arm 

 and the other into a W6 arm, making 10 matched arms of each type. The 

 tests were made on an Amsler testing machine. The average breaking load 

 at the end pinholes was 1159 pounds for the JW arms and 1002 pounds for 

 W6 arms. 



At the same time an estimate was made of the theoretical strength rela- 

 tionship between the two types by means of the moment diagrams shown in 

 Fig. 8. In this figure are shown the graphs of the resisting moments (fiber 

 stress factor — 5000 psi) of clear JW and clear W6 arms, together with the 

 graphs of the bending moments due to the maximum loads these arms would 

 withstand when the loads are concentrated at the end pinholes. These 

 maximum loads were determined by dividing the moments at the points of 

 coincidence between the graphs (critical pole pinhole sections) by the dis- 

 tances to the end pinholes. The maximum loads, so determined, are 608 

 pounds for the JW arm and 532 pounds for the W6 arm. The fact that 

 these loads are low as compared with the actual breaking loads shows, of 

 course, that the average ultimate fiber stress developed by these selected 

 arms was considerably greater than 5000 psi, which is not surprising in view 

 of their exceptionally high quality. However, so far as the information 

 sought is concerned — namely, to determine not the actual strength but the 

 strength relationship between the two types — the result would be the same 

 regardless of the fiber stress factor used in the moment diagram. 



The ratio of the strength of the W6 arm to that of the JW arm as shown 

 both by the actual strength tests and by the moment diagrams was as fol- 

 lows: 



1002 ,, ,^„ 

 Actual strength tests - —r^ X 100 = 



Moment diagrams — ttt^ X 100 = 

 OOo 



Strength Ratio W6 to JW 



(Per cent) 



86.5 



87.5 



These ratios show a remarkably close agreement between theory and ac- 

 tuahty and justify the belief that the crossarm moment diagram may be 

 employed to obtain reasonably accurate estimates of relative bending 

 strength. 



Summary 



The results of this study may be summarized as follows: 

 1. The moment diagram is a useful guide in setting specification 

 limitations on defects. 



