Art.—XLV. An Arrangement for quieting the Flow of a Stream of Disturbed Water.
[Read before the Philosophical Institute of Canterbury, 6th September, 1916; received by Editors, 30th December, 1916; issued separately, 10th December, 1917]
In order to measure with reasonable accuracy the discharge of water over a notch or weir it is necessary to first destroy the waves and eddies present in the stream—in fact, to approximate as closely as possible the conditions of stream-line flow. In practice the following methods have been adopted:—
(1.) The use of a pond above the notch.
(2.) The use of deflector plates.
(3.) The use of perforated plates.
(4.) The use of wire-gauze screens.
The use of a pond above the notch is an efficient and simple method of dealing with the discharge of small streams. The pond is easily constructed; the water is brought nearly to rest and eddies are damped out. It is, however, ineffective in windy weather, and evidently cannot be adopted for flows of any magnitude, nor in the laboratory.
Deflector plates reduce surface-disturbances, but the presence of the plates gives rise to a new series of eddies (fig. 1). For this reason deflector plates are often combined with the perforated plates or wire-gauze screens
With perforated plates the dead-water behind the unperforated portion of the plate gives rise to eddies; surface-disturbance also occurs owing to the drop of a number of small streams (fig 2).
These disturbances are smaller in the case of wire-gauze screens, which, however, are difficult to keep in a sufficiently clean condition to permit of the necessary freedom of flow.
None of these methods being sufficiently effective to quiet the flow to the degree required for some experiments on surface friction, the author devised the following arrangement:—
After being roughly quieted by ordinary methods, the water is passed through a vertical screen (fig. 3) consisting of bars of fish-shaped section, the up-stream end of each being convex and one-third of the total length, the down-stream end being concave and terminating in a knife-edge. There
is thus no dead-water, and the formation of eddies from this source is prevented, whilst eddies previously existent are destroyed by the water being streamed through cross-sections which first gradually decrease and then gradually increase.
One such screen has proved sufficient for the author's purposes, but in some cases it is possible that the addition of a second screen at right angles to the first might be desirable.