46 



Results 



When dye was injected into the boundary layer 

 through the pressure taps along the centerline of 

 the airfoil, it usually did not all move directly 

 upstream on the centerline. Some of the dye moved 

 initially spanwise and then upstream. Regardless 

 of the path of the dye, its motion was never steady. 

 When the airfoil surface was polished and the most 

 accessible of the screens in the settling chamber 

 cleaned, the most forward upstream position of the 

 dye on the unheated airfoil was 4.1 inches (x/L = 

 0.45). With heating, the patter of upstream dye 

 flow remained indistinguishable from the xinheated 

 case. Thus for the wall overheats tested, up to 

 80°F measured at x = 4.01" (x/L = 0.44 in Figure 

 23) , there was no separation delay discernable 

 using the dye injection method. 



In looking at pressure distributions, separation 

 is identified as the point where the experimental 

 pressure distribution departs from the theoretical 

 recompression distribution on the aft portion of 

 the airfoil. The pressure taps will not indicate 

 a separated boundary layer unless there is con- 

 tinuous separation at the tap's position. Thus 

 unless the upstream motion of the dye is very 

 steady, which is usually not the case, the position 

 of separation as determined by the most upstream 

 penetration of dye is consistently farther upstream 

 than indicated by pressure distributions. 



The separation point by examination of pressure 

 distributions on unheated airfoil occurs at x=4.9" 

 (x/L = 9.53) . This is close to the location x = 5" 

 predicted for separation using the Thwaites method. 

 Heating, as reported by Timko and Prahl (1977) , 

 caused no significant alteration in the pressure 



10 20 30 40 50 60 70 80 



T -T °F 

 FIGURE 24. Variation, of shape factor H with wall heating. 



distributions and so again one cannot point to any 

 delay of separation by heating from these data. 



Since the first two indicators showed negligible 

 shift of separation with heating, the boundary layer 

 velocity profiles were measured in some detail at 

 a point upstream of separation with and without 

 heating. Figure 24 shows the results for boundary 

 layer shape factor at a station 3.88" downstream 

 of the leading edge. Heating causes a reduction 

 in shape factor from the unheated value. The un- 

 heated profiles correspond to -0. 17 < B < -0. 15 

 and the slight reduction in shape factor with 

 heating is in accordance with expectation from the 

 similar solutions of Wazzan and Gazley (1977). 

 Despite these shape factor reductions the profiles 

 are changing so rapidly with longitudinal distance 

 (hence the scatter in Figure 24) that the separation 

 location is hardly affected. 



Thus for the amounts of wall heating employed 

 in this study the separation point does not move 

 noticeably from its unheated position. This in a 

 sense confirms the results of Aroesty and Berqer 

 (1975) and of Strazisar (1975) (Figure 20) which 

 show the theoretical insensitivity of the value 

 of S at separation to heating. 



6. CONCLUDING REMARKS 



The studies to date reported herein together with 

 those of Wazzan et al. (1970, 1977), Barker (1978) 

 and others are such as to justify further investi- 

 gation of the various elements of the heating 

 phenomenon. Among the factors affecting the prac- 

 tical application of heating is the combined effect 

 of heating and roughness on stability and transition. 

 The work of Kosecoff, Ko, and Merkle (1976) suggests 

 that the roughness effect is due to the instability 

 of the mean profile as distorted by the roughness. 

 An alternative view being investigated at CWRU is 

 that the roughness introduces disturbances into 

 the boundary layer that may subsequently be ampli- 

 fied by the Tollmien-Schlichting mechanism. In 

 this view the wavelength of the roughness is im- 

 portant as well as its height. An experiment has 

 been planned that will map out the mean and dis- 

 turbance flow-fields in the vicinity of roughness 

 elements so that the relevant mechanism can be 

 identified. This will provide a fluid mechanic 

 characterization of roughness and help in further 

 assessment of the effects of roughness on trans- 

 ition of heated water boundary layers. With further 

 attention given also to heat exchanger design pro- 

 pulsion system, and fabrication techniques, there 

 are promising prospects for the achievement of 

 drag reduction by heating in water. 



ACKNOWLEDGEMENTS 



The author wishes to acknowledge the participation 

 of the following colleagues in the effort reported 

 in this paper: Dr. J. M. Prahl, Dr. M. Nice, 

 Dr. R. L. Lowell, Dr. A. Strazisar, and Mr. M. 

 Timko. All of us are grateful for the sponsorship 

 of the work by the Fluid Dynamics Program of the 

 Office of Naval Research and by the General Hydro- 

 dynamics Research Program of the David W. Taylor 

 Naval Ship Research and Development Center. 



