79 



VIBRATION, 

 ISOLATION 

 SECTION 



UPSTREAM 



BALL 



VALVE 



'1/8 CELL H/SCELL 34 MESH 



HONEYCOMB HONEYCOMB (2 SCREENS) 



3" THICK 3" THICK 



TEST' 

 SECTION 



FIGURE 2. Schematic of turbulence manage- 

 ment system. 



section will tend to "relaminarize, " as described 

 by Laiinder (1964) and Back et al. (1969). However, 

 this would leave us with unknown initial conditons 

 at the entrance to the test section. Therefore we 

 have added the suction section to completely remove 

 the turbulent boundary layer. This section has a 

 0.1 m length of porous wall surrounded by an annular 

 plenum chamber. The suction flow from the plenum 

 is controlled by a valve and a Venturi meter. At 

 each test section velocity above 9 m/sec, the 

 suction flow is adjusted to the minimum value 

 necessary to remove the turbulent boundary layer 

 at the contraction entrance. 



Contraction and Test Section 



The 35:1 contraction was designed by a potential 

 flow calculation using the method of Chmielewski 

 (1974). The length . to diameter ratio of the 

 contraction was chosen by balancing the effect of 

 relaminarization with that of the Goertler insta- 

 bility in the concave-curved portion. A careful 

 study of these two effects led to a length to 

 diameter ratio of 2.25, which made the contraction 

 1.37 m long. The contraction was constructed in 

 two sections: a fiberglass upstream half and an 

 aluminum downstream half. The joint between the 

 two sections is in the region of greatest favorable 

 pressure gradient, and has no measurable step across 

 it. 



Recent velocity measurements in the test section 

 (discussed below) have led to the design and con- 

 struction of a new contraction section to replace 

 the original one. The new contraction will have 

 an annular bleed flow surrounding an entrance 

 section which is all convex. In this way the 

 concave-curved wall, which can produce Goertler 

 vortices, will be avoided entirely. Results using 

 this new contraction will soon be available. 



The flow tube test section is 6.4 m in length 

 and 0.102 m in diameter, with a 2.5 cm wall thickness. 

 It is made of aluminum, and the inside wall has been 

 polished to a surface roughness of less than 10"' m 

 RMS (4 micro-inches) . Surface waviness has been 

 measured as less than one part per thousand for 

 wavelengths less than 2 cm. The tube has been 

 optically aligned on site so that it is straight to 

 within less than 0.018 cm over its entire length. 

 The outside wall is covered with electrical band 

 heaters, which are connected together in groups 

 covering about 0.30 m of length. Each heater group 



is servocontrolled by a system which maintains a 

 preset temperature on a thermocouple located near 

 the inside tube wall. In this way the inside wall 

 temperature can be controlled independently of flow 

 velocity, and different variations of temperature 

 along the tube length can be studied. 



To avoid tripping the boundary layer, no pene- 

 trations of the inside wall are allowed except at 

 the downstream end. The only instrumentation in 

 the test section is an array of thermocouples within 

 the wall, spaced along the tube length. At each 

 location, there is one thermocouple on the outside 

 surface and one in a small hole drilled to within 

 0.15 cm of the inside surface. The temperature 

 difference between the two thermocouples determines 

 the heat flux through the wall at a particular 

 location. Since heat flux increases by a factor of 

 about ten at the transition point, these temperature 

 measurements should provide a good transition 

 indicator. A total of 53 thermocouple voltages are 

 digitized and recorded. 



During the earlier experiments, there was a single 

 hot film anemometer probe at the downstream end of 

 the test section. This probe was located within 

 the boundary layer and was used to indicate inter- 

 mittency only. In the more recent measurements, a 

 new instrumented section has been developed and 

 installed on the- downstream end of the test section. 

 This section is 0.61 m long and its inside diameter 

 matches that of the test section to within 2 x lO"^ 

 m. Two types of measurement can be made in the 

 instrumented section. Very small Pitot tubes can 

 be used to traverse the boundary layer and measure 

 mean velocity profiles, and flush mounted hot films 

 can determine intermittency at various locations. 



Since the boundary layer is typically less than 

 0.5 cm thick, the Pitot tubes must be very small. 

 The one being used at present has a cross-section 

 of 0.013 X 0.076 cm. The smaller dimension is 

 oriented in the direction perpendicular to the wall. 

 The tube is traversed from the wall to the free 

 stream by a micrometer, which can position it with 

 an uncertainty of ±0.002 cm. In addition, the 

 entire central portion of the tube can be rotated 

 in the azimuthal direction so that the Pitot tube 

 can be traversed about the circumference of the 

 test section. The azimuthal rotation can be per- 

 formed while the experiment is running. 



The hot film anemometers in the instrumented 

 section are all mounted flush with the wall to avoid 

 tripping the boundary layer. The Pitot tiobes are 

 removed from the section while hot film measurements 



