A numerical scheme was devised to compute Xp/do> Es y1/b, 
for known values of D/do- The numerical method runs as follows. By 
inspection of Equation A-7, maximum and minimum values of ty which 
render 8 between 90 degrees and 30 degrees were determined. Since 
y; is positive, only positive values of rj must be considered. 
Further, it was determined from Equation A-7 that for 6 to lie 
between 90 degrees and 30 degrees, t, must range between O and 
0.50. A known value of t, yields 6 from Equation A-7. For a 
selected value of o with known tj and 6, parameters D/Dats 1G/Deys 
¥1/bo and xp/bo can be determined from Equations A-6, A-5, A-8 
and A-10 respectively. Thus computing a set of values for Xp/Do 
r/bo and y1/b, for a known D/b,. A series of similar sets were 
computed by increasing ty by 0.05 each time up to a final value 
of 0.50. A total of three series of sets were computed for 
different 6 of 12, 10 and 7.7 respectively. 
For easy usage the parameters 6, y1/d5> E/De and xp/b, 
were plotted against D/do for values of D/b. ranging from 1 
to 1000. These plots are shown in Figures A-2 through A-5. 
It should be added here that flow parameters become independent 
of the parameter D/bg for D/b, greater than 35. Also the analysis 
becomes inaccurate for D/by less than 3. Further, the value 
of o, , the spread parameter chosen can affect the flow parameters 
appreciably. 
One last remark of interest is regarding the value of 
o, the spread parameter to be used while using these curves. 
Because of the curvature effect ao of 7.7 does not apply. 
It is reported in Reference 1 that a value of 12 for o gives 
flow parameters values that are in fair agreement with the 
experimental data. 
28 
