128 



of body length, the measured shear stress and ve- 

 locity profiles became smaller than those predicted 

 by the theory. These differences are more notice- 

 able over the last 5 percent of the body length 

 where the boundary-layer thicknesses are greater 

 than the radii of the bodies, especially for fuller 

 sterns. 



In order to examine the thick stern boundary- 

 layer properties in detail, it is necessary to 

 measure the distributions of static pressure, tur- 

 bulence intensities and Reynolds stress across the 

 thick stern boundary layer. The magnitudes of the 

 eddy viscosity and the mixing-length parameter were 

 determined and compared with those obtained for 

 thin boundary layers. It is found that the eddy 

 viscosity and the mixing length for thick boundary 

 layers are smaller than those of thin boundary 

 layers. An improvement to the Douglas CS differ- 

 ential method can be made by modifying the mixing- 

 length model in the tail region. The distributions 

 of measured static pressure, which were found to be 

 nonuniform across the thick stern boundary layers 

 and near wake, can be approximated very well by 

 potential flow computations for the displacement 

 bodies. The gross curvature effects of the mean 

 streamlines on the static pressure distributions 

 outside the displacement surface are represented 

 very well by those of the potential-flow stream- 



lines of the fictitious displacement body. Thus, 

 the nonuniform static pressure distributions across 

 the thick stern boundary layer can be interpreted 

 mainly as an inviscid phenomenon and can be assumed 

 to have little effect on the stern boundary-layer 

 development. 



Two axisymmetric bodies without flow separation. 

 Afterbodies 1 and 2 of Huang et al. (1976) , were 

 chosen for this investigation. Their geometric 

 simplicity offers considerable experimental and 

 computational convenience in treating fundamental 

 aspects of thick stern boundary layers. Afterbody 

 1 is a fine convex stern while Afterbody 2 is a 

 full convex stern. 



In the following discussion, the experimental 

 techniques and geometries of the model are given 

 in detail. The measurements of mean velocities, 

 turbulence intensities, and Reynolds stresses were 

 analyzed to obtain eddy viscosity and mixing length. 

 The application of the present results to improve- 

 ment of the accuracy of boundary-layer computations 

 over the entire stern is outlined. 



2. WIND TUNNEL AND MODELS 



The experimental investigation was conducted in the 

 wind tunnel of the DTNSRDC anechoic flow facility. 



TABLE 1 - Offsets for Model 1 



X/L 



X/L 



Y/L 



Y/R 



X/L 



Y/R 



