IntroductionT HE hot-wire anemometer has been an important tool for turbulent flow measurement for several decades. Although it is being superceded by optical methods in some instances, there are still many applications for which the hot wire is the simplest and least expensive technique available. For this reason, and because so much data have been published that relies on hot-wire measurements, there is still considerable interest in understanding its behavior.In essence, the hot-wire anemometer measures the rate at which heat is transferred from a heated filament to the flow in which it is immersed. 1 This measurement, however, is not directly useful; fluid dynamicists are more interested in parameters such as root mean square fluctuations of temperature, density, velocity, or pressure, depending on the application. The technique that has been used most commonly to obtain these parameters requires neglect of pressure fluctuations. The inherent uncertainty of this method was calculated as a function of the magnitude of the pressure fluctuations by Johnson and Rose 2 for the case of negligible total temperature fluctuations. However, their analysis does not predict the probable direction of the error, nor does it address the more general case of nonadiabatic flows.The purpose of this Note is to introduce a novel method of analyzing hot-wire measurements that employs two new elements: first, the use of an independent measurement (or estimate) of rms pressure fluctuations; and second, an assumption concerning the special nature of the pressure fluctuation field.A hot-wire probe consists of a very fine metallic wire supported by two conducting prongs. The wire is heated by current passing through it and at the same time cooled by the fluid surrounding it. Analyzing the fluctuations in the voltage across the hot wire to obtain fluctuations in the flow variables is substantially simplified if the following conditions hold. First, the fluctuations of flow variables must be small compared to their time-averaged or ensemble-averaged values. Second, the wire must be oriented normal to the mean flow direction so that there are negligible contributions to the fluctuating signal from the transverse velocity components. As a result of these assumptions, the fluctuating component of the hot-wire voltage may be written 1where E is the hot-wire voltage, p is density, u is streamwise velocity, T t is the total (or stagnation) temperature, and the primes indicate small fluctuating components of the flow variables. The hot-wire sensitivities S p9 S u , and S T/ are defined by logarithmic derivatives, for example S u = [d(?nE)/d(?nu)] pyTt , and must be determined individually for each probe. The sensitivities are also functions of the wire operating temperature and, to some extent, the flow conditions. Note that Eq.(1) applies to both constant-temperature and constant-current modes of hot-wire operation. Hot-wire response may be simplified further if the Mach number is greater than 1.2, in which case 1 S p « S u = S m , where m = ...