radial, and azimuthal directions, respectively. With the probe elements 

 alined vertically, the Reynolds stress and the axial and radial fluc- 

 tuations are computed by an on-line computer with a sampling rating of 



2 

 1024 data values in 8 seconds. To obtain the azimuthal fluctuation w' , 



the hot film sensor is rotated 90 degrees. Data points are accepted at the 



same rate and analyzed by the same computer. 



Figure 9 shows the nondimensionalized measured distributions of 



2 



Reynolds stress -u'v'/U_ and the three components of turbulent fluctuations 



Vu' /U ,Vv' /U„, andyw' /U at various axial locations along the 

 axisymmetric body. The axial fluctuation component is the largest of the 

 three components for all axial locations. The radial velocity fluctuation 

 component has the smallest value. The intensity of each of the three 

 components is larger near the surface of the stern and reduces as the 

 boundary-layer thickens for x/L < 0.95. A sharp gradient exists in the 

 turbulence velocity components near the surface since velocities go to 

 zero at the wall. 



The maximum value of the nondimensionalized Reynolds stress -u'v'/U^, 

 occurs near the wall for x/L < 0.873 with a strong variation in the radial 

 direction.* As the axial location increases and the boundary layer thick- 

 ends, the radial location of maximum Reynolds stress moves away from the 

 wall. The Reynolds stress reduces quickly from the maximum value to zero 

 at the edge of the boundary layer. At axial locations of x/L = 0.951 and 



0.987, the measured values of -u'v reduce from the maximum value to 



zero at the wall. These results are consistent with those of Huang et al. 



for the two afterbody shapes tested. 



Huang et al. also have presented measured distributions of a turbu- 



2 2 2 2 2 



lence structure parameter a where a = -u'v'/q and q = u' + v' + w' . 



Thin boundary layer data show a to have a constant value of approximately 



1 

 0.15 between 0.056 and 0.86. The data of Huang et al. indicated that for 



a thick axisymmetric boundary layer, a^ had a value of approximately 0.16 



*The spatial resolution of the "X" hot-film probe used may not be fine 

 enough to measure the Reynolds stresses precisely near the wall of the thin 

 boundary layer . 



35 



