441 





.ii^^^:^'^^y' 



^^gp??»-' 



'i^Mttt- 



i**^ 



*^?'5y^ =^!^'^^r-X: A 



;^=9o" 



1 mm 



FIGURE 15. Photograph showing boundary layer flow about SST hemispherical nose when a solution of 500 ppm 

 Polyox WSR-301 is injected. The flow is from left to right. V = 4 m/s. 



were made at polymer concentrations of 100, 50, 

 and 20 ppm. The injection rate was such that Vi/Vo 

 = 0.2. The phenomena observed at these lower 

 concentrations were the same as those found at 500 

 ppm. Recently, Gates (1977) studied the influence 

 of polymer additives on laminar flow separation at 

 low injection rates, and was able to find inter- 

 mediate stages of separation suppression. 



The study on the influence of polymer additives 

 on laminar flow separation has been limited so far 

 to the case where the polymer is present only in 

 a thin layer adjacent to the body (the "inner part" 

 of the boundary layer) . To study the influence of 

 polymer additives present in the "outer part" of 

 the boundary layer, additional tests with the SST 

 hemispherical nose were made in which the tunnel 

 was filled with a 50 ppm Polyox WSR-301 solution. 

 To prevent polymer degradation, the water speed in 

 the test section was set at a low value of 4 m/s. 

 Three different solutions were injected: a solution 



0.04 



0.03 



0.02 



* €) « e e c e *•> - 'lJ»*«c«>€»<m)e«) 



%° ° o 





-/-.. 



^^-4^;:;/' 





▼ aii'^iA '^ 



e 0.1 ppm D 10 ppm '^^aJt 



ppm 

 V 20 



• 50 



* 100 

 ■ 200 

 T 500 



J I I I I I 



15 

 Reynolds Number x10~ 



20 



25 



30 



FIGURE 16. Friction factor of Polyox WSR-301 solu- 

 tions in water as a function of Reynolds number, 

 according to Van der Meulen (1974a) . 



of 2 percent NaCl , a solution of 2 percent NaCl + 

 50 ppm Polyox and a solution of 2 percent NaCl + 

 500 ppm Polyox. The injection velocity V^ was 

 0.8 m/s (V.[/Vq = 0.2). Photographs showing the 

 boundary layer flow are presented in Figure 17. 

 For comparison a photograph is included showing 

 the influence of polymer injection when the tunnel 

 is filled with water (Figure 17a) . When a 2 per- 

 cent NaCl solution is injected (Figure 17b) , the 

 boundary layer first shows a tendency to become 

 unstable but further downstream the instabilities 

 are suppressed and the boundary layer is laminar 

 again. When a 2 percent NaCl + 50 ppm Polyox 

 solution is injected (Figure 17c) , the boundary 

 layer first shows a slight tendency to become 

 unstable , but further downstream the boundary layer 

 is laminar. When a 2 percent NaCl + 500 ppm Polyox 

 solution is injected (Figure 17d) , the boundary 

 layer remains completely laminar, till the end of 

 the hologram. The conclusions to be derived from 

 these observations are that the presence of the 

 polymer in the "inner part" of the boundary layer 

 leads to destabilization, whereas the presence of 

 the polymer in the "outer part" of the boundary 

 layer leads to stabilization, and the latter effect 

 is predominant. In all cases considered (Figure 

 17) , laminar flow separation is suppressed. 



An explanation of the various phenomena observed 

 can, as yet, not be given. Apparently, some of the 

 phenomena are in agreement with those reported 

 elsewhere, others may not have been observed before. 

 This is mainly due to the fact that numerous studies 

 have been made on drag reduction in turbulent flow, 

 but only a few were made on the influence of polymer 

 additives on laminar flow. In studying laminar 

 flow around circular cylinders, James and Acosta 

 (1970) found that the streamline patterns with 

 dilute polymer solutions were significantly different 

 from those with Newtonian fluids because of visco- 

 elastic effects. These effects may also play a 

 dominant role in eliminating flow separation in 

 those cases that the boundary layer remains laminar. 

 In those cases where the boundary layer becomes 

 turbulent due to the presence of the polymer in the 

 "inner part" of the boundary layer, it is still 

 questionable whether flow separation is eliminated 

 by early turbulence by viscoelastic effects, or by 

 a combination of these. The occurrence of early 

 turbulence as found in the present study and reported 

 before [Van der Meulen (1976a, 1976b), Gates (1977)] 

 is consistent with the findings of others. According 

 to Lumley (1973) , polymer solutions producing drag 



