Experimental investigation and analysis of the velocity variations in turbulent flow

Abstract

Of the methods used to record turbulent motion in air the use of a hot-wire anemometer has received most attention; with the result the conditions under which a hot wire will accurately record the velocity variations associated with this type of motion are now well known. Many records have been obtained, all of which confirm the irregular character of turbulence, whether produced in the tree stream of the wind tunnel, or in the disturbed region near the boundary of a model. Various attempts have been made from time to time to analyse the results by statistical methods, but partly owing to the difficulty of obtaining lengthy records sufficiently representative of average conditions, and partly because of the labour involved in the process of reduction, no success appears to have been achieved in this direction. Within the last year a new method of analysis suggested by professor G. I. Taylor, which is also capable of discriminating between different types of turbulent motion, has been developed and applied to the analysis of eddying flow in a wind tunnel. The method depends on the use of a hot-wire anemometer for recording velocity variations; but in place of an Einthoven oscillograph a special type of oscillograph having a mirror attached to the suspension fibres is used so that currents produced by the velocity changes are indicated by the horizontal motion of the reflected image of an illuminated slit. If the image falls on a dropping plate in the usual way, it traces the details of the turbulent motion and provides a record in the form of a black line on a light background; that is, the reverse of that given by the Einthoven oscillograph. This all-important difference gives the new method a power of analysis that the older one lacks; for, if instead of moving across a dropping plate the reflected image moves across a stationary plats, tracing and retracing its path in accordance with the imposed changes of speed, the plate on being developed displays a dark band of variable density representing tbs aggregate effect of every velocity change occurring during an experiment, which may last as long as 5 or 10 minutes. It thus constitutes a detailed record whose density at any point determined photometrically measures the total time of exposure at that point. such information enables one to calculate the relative number of times a given displacement has been reached or exceeded in the course of an experiment. What is more important, when used in conjunction with the known relationship between the displacement of the image and wind speed, it provides sufficient data from which frequency curves can be plotted indicating the actual distribution of the recorded velocity variations.