under investigation.* In this instance, if /^ is the mean square of the response to the original 

 turbulence, /^^ j, the response with the addition of the superimposed field, the signal from the 

 superimposed field alone is 



Now since the turbulence corresponding to /^ is known, the turbulence of the stream may be 

 readily obtained without knowing the wire resistance or any of the constants of the amplifier 

 or metering circuits. More important, the accumulation of dirt and surface film on the wire 

 would be unimportant provided the time response of the wire were not seriously impaired. 



Mr. Stracke worked for some time on a dynamic calibration technique in which the wire 

 is given a known vibratory motion. The probe was mounted on a specially designed arm which 

 could be mechanically vibrated. The wire moved in a small arc which was essentially parallel 

 to the flow. There are a number of inherent difficulties in this vibration method which were 

 never completely ironed out. First, it is necessary that the vibrator be well designed to elim- 

 inate spurious motions so that the displacements and amplitudes can be accurately measured. 

 Corrections must be made for a possible sag in the wire as it moves. As strains in the wire 

 would become excessive if the vibration frequency were too high, it is necessary to limit the 

 frequency to about 10 cycles per second. It is also necessary for the wire, the amplifier, and 

 associated circuits to have a flat frequency response over the turbulence range down to the 

 vibration frequency of the mechanical oscillator. Some of these difficulties may be overcome 

 by comparing the vibration response to the wire response in a known turbulence field, such as 

 that behind a grid or rod. Details of the technique must still be worked out. 



Another method of dynamic calibration which has been suggested is to insert a grid or 

 cylinder a known distance ahead of the wire. Since the turbulence of the wake of these objects 

 is already known a calibration could be obtained as before, provided the two turbulence fields 

 are not correlated. The grid or cylinder could be swung in and out of position without disturb- 

 ing the wire. The principle objection to this method of calibration is that it would be difficult 

 to use in a confined space. 



INSTRUMENTATION 



Very little work has been done in developing instrumentation for use with the hot wire at 

 the Taylor Model Basin since the early constant-current instruments were procured by Macov- 

 sky and Stracke.^ Although the recent studies were made with a constant-current a-c bridge 

 it would be highly desirable to use a constant-temperature circuit for turbulence measurements. 

 The constant-temperature wire has advantages even though the frequencies of the turbulence 

 fluctuations expected to exist in water are low enough that compensation would not be a ser- 

 ious problem. In flows where the velocity remains constant for only short intervals it would 



*An investigation of how two such turbulent fields are correlated will be made in the TMB Low-Turbulence 

 Wind Tunnel. 



