97 



FIGURE 24. Flow past the hemisphere body with injec- 

 tion of 100 wppm Polyox at a Reynolds number of 3.9 x 

 10^. The dimensionless injection values are: (a) G 

 0, no injection, (b) 0.5 x 10"6^ (<,) i.i > 10"^, 

 (d) 1.7 X 10"6, (e) 2.9 



X 10"6, 

 10"5. 



material by the mass flux of the boundary layer 

 displacement flow. Although this is an arbitrary 

 normalization, in the present experiments the dis- 

 placement effect of the injectant fluid was always 

 much less than the boundary layer displacement 

 thickness, 6^. Thus we define a quantity G 



cQ 



ttDV 6 



to tS 



where c is the polymer concentration (weight basis) 

 in the injectant Q the volume flow rate of injectant 

 (basically the same fluid as the test medium) with 

 D and Vcc being the body diameter and tunnel ve- 

 locity respectively. For the NSRDC and hemisphere 

 models 6^3 was calculated at the position of the 

 laminar separation whereas for the Schiebe body it 

 was arbitrarily calculated at S/D = 1.00. The pres- 

 ent results, so normalized, are presented in Fig- 

 ures 26, 27, and 28. As with the freestream 

 turbulence level, no change in the position of 

 separation on the NSRDC and hemisphere nose models 

 was observed when polymer was injected into the 

 boundary layer . 



The results of the experiments show the presence 

 of very small quantities of Polyox to be destabiliz- 



ing to the laminar boundary layers on the present 

 test models. This desteibilization effect has been 

 observed before: in fully developed cavity flows 

 past spheres and cylinders Brennen (1970) observed 

 distortions in the cavity surface and separation 

 line due to the presence of polymer. Brennan at- 

 tributed the changes in cavity appearance to a 

 polymer induced instability in the wetted surface 

 flow on the headform. Sarpkaya (1973, 1974) in- 

 vestigated the flow of dilute polymer solutions 

 about cylinders and several airfoils and also ex- 

 plained his observations by suggesting a polymer 

 induced instability in the laminar boundary layer. 

 Some later experiments by Tagori et al . (1974) 

 support some of Sarpkaya 's speculation for one of 

 the airfoils. 



A destabilizing effect is rather contrary to the 

 general impression obtained from the available lit- 

 erature on the effects of drag-reducing polymers on 

 fluid friction [see for example Hoyt (1972)]. We 

 were unable however, to find in the available lit- 

 erature any satisfactory explanation of the effect 

 on transition of the polymer fluids. 



FIGURE 25. Flow past the Schiebe body at a Reynolds 

 number of 4.2 x 10^ with injection of 500 wppm Polyox. 

 The dimensionless injection parameters are, (a) G = 0, 

 (b) 2.3 X 10"6, (c) 1.5 X 10-5, (^) 2.9 X 10-5. J-3CJ, 

 frame is 0.2 body diameters in length and they are 

 centered at arc length ratios of 0.82, 0.75, 0.6, 

 0.53, respectively. 



