80 



are being made. The films are used only to determine 

 intermittency , hence they are not calibrated. There 

 are eight hot film locations — two streamwise sep- 

 arated stations each having four probes at different 

 azimuthal angles. All eight outputs can be displayed 

 simultaneously on oscilloscope traces or recorded 

 on a photographic strip-chart recorder. 



A high static pressure must be maintained in the 

 test section to avoid possible cavitation or out- 

 gassing from heated walls. Therefore the pressure 

 loss for controlling the flow velocity is located 

 at the downstream end of the experiment. Originally, 

 a set of sharp-edged orifice plates was used on the 

 end of a 1 m long extension tube added to the test 

 section. Concern over possible upstream influence 

 of the disturbances generated at the orifice plate 

 led to the development of a smooth contraction 

 section for the downstream end. With the smooth 

 contraction, it is possible to maintain laminar 

 flow all the way to the exit of the experiment, and 

 thus determine transition by flow visualization in 

 the exit jet. In addition, a "plug nozzle" has 

 been developed, which consists of a strut-supported 

 central cone which can be moved in and out of the 

 end of the test section. This adjustable exit 

 valve permits us to vary the test section static 

 pressure independently of flow velocity while main- 

 taining laminar flow all the way to the exit. With 

 any of these possible exit conditions, the test 

 section velocity can be determined from the test 

 section static pressure and the known discharge 

 coefficient of the nozzle. 



3 . RESULTS 



Free Stream Turbulence 



Mean and fluctuating velocities were measured in 

 the settling chamber by a cylindrical hot film 

 anemometer. The probe penetrated the settling 

 chamber wall 0.1 m downstream of the boundary layer 

 suction section, and could be traversed from the 

 wall to the centerline. Mean velocities and tur- 

 bulence levels were measured at many points, and 

 turbulence spectra were measured at two or three 

 points for each flow condition. In addition, a 

 1.2 m long instrumented straight tube could be 

 substituted for the 6.4 m test section. This short 

 tube contained a Pitot tube, accelerometers , and 

 hot film probes . The unheated transition Reynolds 

 number was measured in the 1.2 m tube for each 

 settling chamber configuration. This Reynolds 

 number varied from 800,000 for the empty settling 

 chamber with no turbulence manipulators to 5.0 x 

 10 for the "best" configuration. This configura- 

 tion (shown in Figure 2) includes one piece of porous 

 foam, two sections of honeycomb, and four screens. 

 The last screen is located 0.3m upstream of the 

 beginning of the contraction, and has a mesh of 24 

 per cm. All screens in the settlinig chamber have 

 more than 55 percent open area, in accordance with 

 the findings of Bradshaw (1965) . 



Detailed results of the velocity measurements 

 in the settling chamber have been reported separately 

 [Barker (1978)], and are only summarized here. At 

 test section velocities less than 9 m/sec, the 

 settling chamber boundary layer remains laminar and 

 the only effect of the suction is to make it thinner. 

 The turbulence level is about 0.07 percent at all 

 distances from the wall for the configuration of 



Figure 2. At higher velocities the turbulence level 

 near the wall reaches 3 or 4 percent with no suction, 

 but remains 0.07 percent at distances from the wall 

 greater than 2 cm. As the suction flow rate is 

 increased, the mean velocity profile shows thinning 

 of the boundary layer and the turbulence level near 

 the wall drops rapidly. At the optimum suction 

 rate, the highest turbulence level near the wall 

 in the settling chamber is about 0.4 percent. The 

 suction has no measurable effect upon the mean 

 velocity profile or turbulence level more than 2 

 cm from the wall. 



The settling chamber velocity measurements and 

 the unheated transition Reynolds numbers indicate 

 that the turbulence management system is performing 

 well. If the turbulence level reduction through 

 the contraction is proportional to the square root 

 of the contraction ratio [Pankhurst and Holder 

 (1952)], then the turbulence level in the test 

 section should be about 0.01 percent. This is 

 lower than the turbulence level recorded in most 

 wind tunnels , and certainly lower than any previ- 

 ously reported water tunnel. 



Transition Reynolds Numbers 



Figure 3 shows measured transition Reynolds numbers 

 as a function of wall overheat for the uniform wall 

 temperature case. The results on the upper curve 

 were obtained with the smooth, laminar flow nozzle 

 at the downstream end of the test section, using 

 flow visualization at the exit to determine tran- 

 sition. The water temperature was approximately 

 50°F (10°C) during these tests. Note that the 

 transition Reynolds number rapidly increases with 

 wall temperature up to 10 °F (6°C) wall overheat, 

 at which it has reached a value of 42 x 10 . 

 This represents a factor of four increase in tran- 

 sition Reynolds number for a relatively small heat 

 input. However, above 10°F there are no further 

 increases in transition Reynolds number, while the 

 theory predicts that it should increase up to about 

 60°F (33°C) overheat. Previously published results 

 [Barker and Jennings (1977) ] have shown that varying 

 the wall temperature distribution does not change 



/ SMOOTH 



WAZZAN, ET. AL. 

 /J= 0.07 



ONE EXTENSION 



6 10 16 20 25 



OVERHEAT, AT (°F) 



FIGURE 3. Transition Reynolds numbers measured at 

 exit: one extension tube. 



