removable turbulence-sensing elements were mounted at various positions within the boundary 

 layer on the forebody of the model. Although no information was obtained as to the intensity 

 or scale of turbulence, the oscillogram records clearly delineated the regions of laminar, trans- 

 itional, and turbulent flow. Typical oscillogram records are shown in Figure 2. 



The same qualitative technique has also been used to determine the spanwise phase 

 configuration of vortex shedding behind a cylinder towed horizontally through the water. Fol- 

 lowing a procedure used by Roshko^ one element was placed at a fixed point behind the cylin- 

 der while a second one traversed the wake along the same horizontal line parallel to the cylin- 

 der. 



RECENT EXPERIMENTAL INVESTIGATIONS TO IMPROVE WIRE STABILITY 



The inherent instabilities of the hot wire as it was developed by Macovsky and Stracke 

 precluded its extensive use for quantitative turbulence measurements in water. Early in 1954, 

 however, interest in obtaining a practical instrument for quantitative turbulence measurements 

 was revived and work on the hot wire was resumed at TMB and also at the Iowa Institute of 

 Hydraulic Research.^ Efforts at both laboratories were directed toward obtaining a hot-wire 

 instrument which would be as easy to operate in water as in air. 



It was discovered independently at both laboratories that much of the wire instability 

 could be eliminated by the use of an alternating heating current. The bubbles which formed 

 on the wire when it was heated with a direct current were largely a result of electrolytic dis- 

 sociation of the water. As the bubbles formed and broke away erratic changes were produced 

 in the rate at which the wire was cooled by the flow. With an alternating heating current of 

 several thousand cycles per second, however, the bubble formation and the resulting resis- 

 tance instability were eliminated. Even in highly aerated water bubble formation was no prob- 

 lem. 



Even with the elimination of bubble formation enough instability remained to make cali- 

 brations uncertain. In order to track down these other instabilities the hot wire was mounted 

 in an a-c bridge circuit and resistance changes could be observed as the bridge became un- 

 balanced. A diagram of the circuit is shown in Figure 3. The heating current was supplied 

 by a power oscillator and the bridge unbalance was observed on an oscillograph. It was nec- 

 essary to use isolation transformers on the input and output of the bridge to avoid ground loops. 



In the preliminary experiments the sensing element was fabricated of 0.3-mil tungsten 

 wire using the same techniques previously developed by Macovsky and Stracke. After all 

 ground loops and electrical pickup had been eliminated an instability persisted which was 

 finally attributed to a galvanic action between the dissimilar metals used in the probe sup- 

 ports, solder, and plating on the wires. Consequently, the old technique of soldering copper- 

 plated tungsten wires to the probe tips was abandoned and unplated tungsten wires were weld- 

 ed directly to the probe holders. The joints and probe holders were painted with a good 



