Some Response Characteristics of Parabolic Hot Films in Water

Abstract

TURBULENCE in liquids presents an important problem for many disciplines from chemical engineering to ocean physics, hydraulic engineering, and fluvial geomorphology. In such areas mean velocities are usually low (less than 5 msec) and spectra of turbulence range over frequencies from about 300 Hz down to a scale determined by that of the experiment. The only instrument presently capable of measuring such a range in a nonlaboratory environment is the hot-film anemometer. The classic field experiment of Grant, Stewart, and Moilliet used a conical hot film probe to test Kolmogorov's hypotheses, while more recent field measurements in hydraulics have been carried out by Holley. The response of such an instrument is a subject of considerable concern, for no adequate theoretical model has been developed for its heat transfer characteristics. The frequency response of hot wires has been extensively studied since King's pioneering work (Davies and Fisher, Champagne et al., and Sandborn. Early work on hot films by Grant et al. and Evans tested the frequency response of conical hot films and found the response flat up to 200 Hz, then showing a tendency to rise with increasing frequency up to 1 kHz. However, the work of Fabula showed the response to drop for glass cylinders, the drop being greater for the 5 mil than 2-mil-diam sensor. No other work appears to have been carried out at frequencies above 1QO Hz. Hot films have come under increasing scrutiny since the commencement of commercial manufacture of the probes. Liu examined the response of a cylindrical hot film over the range 2-38 Hz and concluded that the response was essentially flat. More recent work of a similar nature was carried out by Rodriguez et al.,' in which probes of several different configurations were vibrated over a frequency range of 5-100 Hz in turbulent pipe flow, with a velocity range of 2-5 msec in viscoelastic solutions. Responses were practically flat over the entire range but showed a tendency to rise at frequencies below 10 Hz which, it was claimed, was due to experimental error. Theoretical studies of the dynamic characteristics of hot films have been carried out by Rasmussen and Bell-