An anemometry technique for turbulence measurements at low velocities

Abstract

A method using symmetrically bent V-shaped hot-wires has been developed for the accurate measurement of low-speed turbulence. Directional characteristics of V-shaped hot-wires at low velocities are investigated, and a generalized expression is derived for the effective cooling velocity. The measurement with V-shaped hot-wires in a pseudo turbulent field, which is artificially produced by shaking the hot-wires with an accurately known motion in a steady flow, has confirmed that the expression for the effective cooling velocity is also valid for instantaneous velocity fluctuations. The accuracy of a practical technique comprising two V-shaped hotwires in an X arrangement is investigated by an error analysis in simulated Gaussian velocity fields using a digital computer.

Evaluation of X-probe response to wire separation for wall turbulence measurements

“…Some comments on the assumptions (i) and (ii), however, will be given later. The assumption (v) was used by Tsuji and Nagano (1989) in the light of the assumption by Bremhorst (1972) to make the spatial resolution analysis possible. From Eq.…”

“…The number of Gaussian random numbers was 100,000 for each velocity component, which was large enough to evaluate the turbulence statistics (Kawall et al 1983;Tsuji and Nagano 1989). It should be recalled here that the effective cooling velocity Ue is consistently positive and thus the basic equation for the hot-wire signal, Eq.…”

“…However, there is only little work on the influence of the differences between the velocity fields at the z-locations of the two wires of an X-probe upon measurement of turbulent statistics (Wyngaard 1968;Bremhorst 1972;Browne et al 1988;Suzuki and Kasagi 1990). Tsuji and Nagano (1989) also investigated measurement errors of an X-probe in conjunction with V-shaped hot-wire anemometry (Hishida and Nagano 1988 a, b) in cases where an effect of spatial difference in an instantaneous velocity field is small. When a scale of turbulence becomes small, as in nearwall turbulence, or a hydrodynamic disturbance caused by probe supports is no longer negligible, the difference in in- Configuration and orientation of an X-probe with respect to flow; a conventional X-probe, b modified X-probe stantaneous velocities sensed by hot wires of an X-probe often become fairly large.…”

“…We use pseudoturbulence with multivariable Gaussian random numbers generated on a computer to evaluate the X-probe response. This method has been successfully used to study cross-wire anemometry in high-intensity turbulence by Kawall et al (1983), Swaminathan et al (1986), and Tsuji and Nagano (1989). In addition, measurement of pipe flow turbulence has been made both with a conventional X-probe and with a modified one, and the results are compared with those obtained by the present error analysis, through which a guide to reliable turbulence measurement with an X-probe is presented.…”