20 



height not exceeding 0.01 -in. were found to be directly in the line of flow 

 of the "peninsula" of turbulence shown in Figure 14a. The surface was care- 

 fully rubbed down and refinished and the entire test repeated. The repeat 

 test gave patterns of which Figure 14b is typical. Dryden has reported a sim- 

 ilar phenomenon. He found that a thin layer of dust on a flat plate tested in 

 a wind tunnel was sufficient to cause an anomalous boundary-layer flow-pattern. 

 These results clearly show that in order to obtain repeatable data the surface 

 of the model must be kept extremely smooth and clean. 



Condition 2 - Stimulation by Rods 



Three surveys were made with a l/8-inch-diameter circular cylinder 

 or rod placed vertically in the longitudinal £ plane of the model at distances 

 forward of the stem of 24, 48, and 192 rod diameters (3, 6, and 24 inches). 

 They produced the distributions of boundary layer flows shown in Figure 15. 

 From these sketches it is seen that at a distance of 48 diameters the rod stim- 

 ulates the boundary layer more effectively than at the other positions. It is 

 interesting to note that, with the rod 192 diameters forward, at speeds great- 

 er than 0.8 knot the shapes of the laminar and transitional areas are different 

 from those found for the other rod settings, indicating stronger stimulation in 

 the neighborhood of the forefoot than at the water surface. From this it 

 would appear that the turbulence produced by the rod decays more rapidly at 

 the surface than at keel depth. However, since no information on such phe- 

 nomena is believed to be available at this time no authoritative statement 

 can be made. It is expected that quantitative measurements of the decay of 

 turbulence behind rods which are to be made at the Taylor Model Basin will 

 provide sufficient data to determine the best position of the rod for adequate 

 stimulation. 



Condition 3 - Stimulation by Trip Wire 



Results of the boundary-layer survey made with the 0.032-inch trip 

 wire in place are summarized in Figure l6. The rapid change in the position 

 of transition behind the trip wire agrees with the result obtained from meas- 

 urements in a water tunnel on two bodies of revolution. 10 A plot of the var- 

 iation of the minimum distance of transition from the stem with wire Reynolds 

 number from the tanker test is given in Figure 17 together with curves deduced 

 from the data given in Figure 13 of Reference 10. Both sets of data exhibit 

 similar change in the position of transition with wire Reynolds number. The 

 local wire Reynolds number increases with depth on the ship model because of 



