96 



TABLE II Approximate Critical Boundary Layer Frequencies 

 for Several Bodies 



4. 17 X 10- 



3350 



2100 



1200 



ments of van der Meulen (1976) in Figure 22. There 

 is good agreement of the experimental results and 

 also with Wazzan's e calculations. 



Polymer Injection 



ObseTva tions 



The influence of gradually increasing the injection 

 rate of the polymer solution upon the basic flow 

 about each test body is illustrated in the sequences 

 of schlieren photographs in Figures 23 through 25. 

 In Figure 23(a) the maximum height of the separa- 

 tion bubble is 0.5mm and on the hemisphere nose 

 body in 24(a) the maximum bubble height is 0.25mm. 

 Unlike the freestream turbulence level, the presence 

 of polymer in the boundary layer was found to 

 influence the basic viscous flow on all the test 

 models. As can be seen in the schlieren photographs, 

 as the polymer injection rate was increased the 

 position of transition moved upstream in each case. 

 For the NSRDC and hemisphere nose models a critical 

 injection rate was reached at which the positions of 

 transition and separation coincided and the laminar 

 separation was eliminated. At injection rates above 

 this critical value the position of transition ap- 



O 0.8 



!= 0.6 



to 



o 



Q. 



"T" 



"T" 



o van der MEULEN (1976) 



• PRESENT OBSERVATIONS 

 (0.05 PERCENT 

 TURBULENCE LEVEL) 



• CALCULATIONS NO 

 HEATING 



• CALCULATIONS I0°F 

 HEATING (WA2ZAN a 

 GAZLEY, 1978) 



I X 10' 2x10^ 3x10* 



BODY REYNOLDS NUMBER-UD/i/ 



4X10' 



FIGURE 22. Comparison of transition observations on 

 the Schiebe body. 



FIGURE 23. Flow past the NSRDC body with injection 

 of 500 wppm Polyox (WSR 301) at a Reynolds number of 

 1.6 X 10^: (a) no injection; (b) 0.1 mJl/sec, G = 

 0.5 X 10~6; (c) 0.3 mH/sec, G = 1.5 x 10"^; (d) 0.5 

 mS,/sec, G = 2.5 X 10"^. G is the dimensionless 

 polymer injection rate. 



peared to move further upstream, but with the limited 

 resolution of the present schlieren system, these 

 poisitions could not be accurately determined. 



Discussion 



It would seem desirable to normalize somehow the 

 injection rate of polymer fluid. We have chosen 

 to do this by dividing the mass flux of polymer 



