EARLY DEVELOPMENT OF THE HOT WIRE FOR 

 USE IN WATER AND ITS APPLICATION 



Beginning in 1946 Macovsky and Stracke developed a constant-current, hot-wire, turbu- 

 lence-sensing element for use in water. ^'-^ The element was heated with a direct current and 

 was used in a conventional constant-current hot-wire circuit which had been developed at the 

 National Bureau of Standards.^ Most of the effort at the Taylor Model Basin was expended in 

 selecting wire material and in developing methods of fabricating the probe and a technique for 

 making turbulence measurements. In addition, the hot-wire element, as originally developed, 

 has been used with varying degrees of success in quantitative turbulence measurements and 

 has been used very successfully for qualitative measurements and for turbulence detection. 

 The choice of a wire material depends not only on favorable electrical properties but 

 also on its ability to withstand the corrosive action of water and the relatively large hydro- 

 dynamic forces. The most suitable material was determined to be tungsten. Probes fabricated 

 from 0.3-mil tungsten wire, about % in. long, had a resistance of from 5 to 10 ohms and could 

 be given a temperature elevation of 15F with a heating current of about 120 milliamperes. 

 Such wires did not corrode and were able to withstand water speeds up to 20 knots without 

 failing. 



In fabricating the probe the wires were carefully copper plated, leaving a gap of the 

 desired length for the sensitive portion. Then the plated portions were soldered across the 

 tips of the metal supports. Finally, the supports were painted with insulating material to re- 

 duce electrolysis. Details of the fabrication technique are reported in References 1 and 2. 

 The major obstacle in obtaining accurate turbulence measurements with a hot wire is 

 the instability of the wire resistance caused by the accumulation of gas bubbles, dirt, and a 

 surface film on the wire. This contamination starts as soon as the wire comes into contact 

 with the water. If the resistance fluctuations arising from changes in the convective cooling 

 of the flow are to be measured quantitatively, all measurements must be made with a freshly 

 cleaned wire a few seconds after it has been placed in the flow. The measuring technique 

 developed by Macovsky and Stracke consisted of removing the wire before each measurement, 

 cleaning it with a camel's hair brush which had been dipped in acid, positioning the probe 

 again, and quickly taking a reading. The heating current in the wire was automatically turned 

 on and off by means of a microswitch as the probe was swung into and out of the water. To 

 check a reading it was necessary to repeat the whole process. Thus the reliability of the 

 turbulence measurements depended a great deal on the patience and dexterity of the investi- 

 gator. With care, measurements could be repeated. 



The foregoing technique was used in the 1/22-scale model of the Taylor Model Basin 

 circulating water channel^ to repeat some of the classic wind-tunnel experiments of free-stream 

 turbulence behind grids and cylinders. In the first experiment the hot-wire sensing element 

 was mounted at various distances downstream from a grid and quantitative measurements of 



References are listed on page 14. 



