TELEPHONE LINE WIRE SPACING PROBLEMS 205 



To simplify to some extent the long process of correlation and 

 analysis, first, the velocities recorded at 16 directions to the line were 

 converted to normal velocities, which should be borne in mind in 

 considering the results. This conversion was made by multiplying 

 the recorded velocities by the cosine of the angle between the normal 

 to the line and the true direction of the wind. The propriety of taking 

 this step was based on appropriate wind tunnel tests made at New 

 York University and reported in the Bell System Technical Journal? 

 From this point on the data were classified according to the procedure 

 outlined in Fig. 7. 



This procedure provided a history of the contacting on each pair 

 arrangement tested. This history detailed the number of contacts 

 occurring in each five-mile-per-hour cell of maximum and five-minute 

 average normal wind velocity for each division of sag. Each division 

 of sag comprised a cell of one to three inches depending upon the 

 length of span. 



These results were analyzed with the view of determining first, the 

 instantaneous velocity, at which contacting begins to occur for each 

 pair arrangement, and second, any relationships existing between the 

 fundamental factors of spacing, sag, span length and such instan- 

 taneous velocities which are hereinafter referred to as threshold 

 velocities. The term "threshold velocity" as used in this article does 

 not relate to the five-minute average velocities but to the maximum 

 velocities previously mentioned. To express threshold velocities in 

 terms of five-minute average velocities it is necessary to divide by 

 the factor of 1.4, referred to above. 



The analysis of the data directed towards determining the first 

 objective or the threshold velocity for each pair arrangement revealed 

 considerable variation in the magnitudes of these velocities for any 

 particular sag. This has been ascribed to the variability of natural 

 winds. Under these conditions it appeared to be appropriate to select 

 the velocity most frequently associated with the beginning of con- 

 tacting as the threshold velocity for a given arrangement at the 

 prevailing sag. Thus the modal value was taken as the threshold 

 velocity, or the nearest velocity when expressed in multiples of five 

 miles per hour to the true modal value. The accompanying Table I 

 lists the threshold velocities for the arrangements tested. 



Regarding the second objective, namely, the analysis of the data 

 for the purpose of determining relationships existing between the 

 fundamental factors of spacing, span length, sag and wind velocity, 



^ "Forces of Oblique Winds on Telephone Wires," J. A. Carr, Bell System Technical 

 Journal, October 1936. 



