ANEMOMETRY, the measurement of air speed. It is difficult to design instruments to measure accurately the speed of moving air; and although recent research has helped, difficulties still persist, particularly as regards the measurement of low air speeds. The trouble is mainly because direct methods are im practicable, so it is necessary to measure some physical effect arising from the motion. Three such effects have been found suitable—namely, pressure changes associated with the motion; mechanical effects, such as the rotation of certain types of wind mills appropriately mounted in the stream; and, lastly, the rate of cooling of a hot body, such as an electrically heated wire, exposed to the current. The first is most important, since a properly de signed instrument, suitably inserted in the stream, experiences a pressure, commonly called the velocity head, entirely char acteristic of the motion, and measurable on a pressure gauge. If such an instrument is constructed on what are now established principles it may be used without calibration as a standard f or measurement of wind speed. This is not true of anemometers which depend on mechanical or electrical effects; these are subject to individual variations and require calibration against a standard instrument of the pressure-tube type.
Pressure-tube anemometers in general consist of two parts ; one is common to all types—namely, an open-ended tube facing the air stream. If flow through this tube is prevented, by connecting the other end to a pressure gauge, the pressure at the mouth is found to be equal invariably to the sum of the static pressure and the velocity head in the stream at that point. The other part of the anemometer varies in detail with different types, but always consists of a second tube with orifices so arranged as to measure either the static pressure in the stream or the static pressure multi plied by a certain coefficient characteristic of the instrument.
We may consider first the case in which the static tube meas ures accurately the static pressure itself. If then the two tubes of the complete anemometer are connected to opposite sides of the same differential pressure gauge, the static pressure portion of the pressure transmitted by the facing or "total head" tube will be balanced by the pressure transmitted from the static tube, and the resultant reading on the gauge will be the velocity head p, which is related to the speed of the air-stream V by the equation where p is measured as force per unit area and p is the air density (mass per unit volume), calculable in any given case from know ledge of the temperature and pressure of the air.
An instrument, whose differential readings measure the velocity head to an accuracy of about o• i % is shown in fig. 1. The two tubes are here arranged concentrically, the inner, or Pitot tube (so called of ter Pitot, who was the first to suggest the use of pressure tube anemometers), measuring the sum of the velocity and static heads—that is the "total" head, whilst the outer one measures the static pressure at the series of small orifices around its periphery as shown on the diagram. Such an instrument forms a convenient standard to which the readings of other types of anemometers may be referred.
In pressure-tube instruments of the second type, in which the static side does not indicate the true static pressure, the differen tial pressure pi is usually given by an equation of the form where k is now a coefficient whose magnitude is in general greater than o.5 and may vary with the wind speed. In the Dines pressure tube, much used in meteorological work, the total head tube is mounted so that it is free to rotate about a vertical axis, and is equipped with a weathercock-vane which always points the tube into the wind. A vertical tube, pierced with a number of small holes past which the wind streams, forms the static side, the con stant k in equation (2) for this instrument being about 0.75. The pressures are usually transmitted to a recording float-manometer, in which the float moves vertically by an amount corresponding to the applied pressure difference.
The motion of the float is com municated to a pen which marks on a chart attached to the surface of a cylindrical drum driven by a clock, and making as a rule one complete revolution in 24 hours.
Continuous records of wind speed extending over this period are thus obtained on one chart.
Many mechanical devices for measuring air speeds have been proposed, but only the Robinson cup anemometer and the vane anemometer now survive to any marked degree. The former (see fig. 2) consists essentially of four hemispherical cups carried with their bases vertical, at the outer ends of four light arms forming a 90° cross in a horizontal plane, and attached to a central sleeve free to rotate about a vertical axis. Opposite cups are arranged so that the concave side of one member of the pair is presented to the air current at the same time as the convex side of the other. Hence, since the aero dynamic force on a cup with its concave face presented to the wind is greater than when the wind is blowing on its convex face, rotation will ensue at a rate depending on the wind speed. The cups are connected by gearing to an indicating mechanism which may either show revolutions, or, as is almost universal practice, units of distance travelled by the wind. In the latter case the distance travelled by the wind for one revolution of the cups is deduced from experience with other instruments of similar type, and the gearing ratio is adjusted to suit this relationship. Under steady conditions, then, the wind speed is obtained by observing, with a stop watch, the motion of the indicating mechanism in a given time.
The ratio of wind speed to linear speed of the centres of the cups is termed the factor of the anemometer. Experimental evi dence indicates that the value of this factor depends on the di mensions of the instrument and also on the wind speed, so that even for any particular instrument the factor is not constant at all speeds. These anemometers must be calibrated in winds of known speeds for accurate work. From tests on many such instruments of different sizes it appears that the factor may vary from about 2.2 at the high wind speeds to possibly a little over three at low speeds. A convenient form of cup anemometer has an electric generator attached to the cup shaft and driven by it. The E.M.F. generated thus depends on the rate of rotation of the cups, and can therefore be used as a measure of the wind speed, the relation between these two quantities being determined by test. An indi cating voltmeter, connected to the generator, can then be cali brated to read wind speed directly on a scale. Alternatively, by the use of a recording voltmeter, continuous records of wind speed can be obtained. This form of instrument gives an instan taneous reading of wind speed, without the necessity for a stop watch, and also readings can be taken at a distance from the anemometer cups themselves. Obviously, the latter feature is of ten one of considerable importance, particularly in meteoro logical work, which is the main field of application of the cup anemometer, since its readings are not affected by changes of wind direction in the horizontal plane. Other forms of distant recording mechanism have also been devised.
The vane anemometer consists of light flat vanes attached to radial arms mounted on a common spindle which rotates in two jewelled bearings. Eight vanes are usually employed; they are inclined at an angle to the axis of the spindle, which is set along the wind direction when the instrument is in use, and the wind forces acting on the vanes cause the spindle to rotate at a rate depending, in any given case, on the air speed. As in the cup anemometer, the motion of the spindle is communicated by a gear train to a pointer, or, more usually, to a number of pointers moving over dials graduated directly in feet. Experience has taught the makers how to proportion the various dimensions and gear ratios so that the number of feet indicated on the dials in unit time is approximately equal to the distance traversed by the wind in the same time, i.e., to the wind speed. Instruments of this type are mainly of value for measuring air speeds in large ducts or ventilating shafts. They are particularly useful for measuring low speeds where the pressures to be measured with pressure-tube instruments are so small as to necessitate the use of extremely sensitive manometers.
Vane anemometers of normal design can be used over a range of wind speeds from about five to so feet per second. For higher speeds, up to ioo feet per second, special instruments are made which are either of heavier construction or, alternatively, incor porate a device by means of which the quantity of air actually passing through the vanes can be reduced. Vane anemometers of specially light construction can also be obtained to measure speeds of less than five feet per second, the lower limit of speed that will cause steady rotation of the vanes being the order of 0.5 feet per second. All instruments need calibration in a wind whose speed can be controlled and adjusted to known values, for it is not possible to predict accurately the readings of a mechanical anemometer from a knowledge of its dimensions only.
In the absence of mechanical friction due to weight of moving parts and to gearing, readings of both cup and vane anemometers would depend only on the wind speed, and so would be independ ent of changes of air density. In most cases these frictional couples will be small compared with the aerodynamic driving torque, and no attention need therefore normally be paid to changes in air density. If, however, the changes of density are large—more than about 5%—the readings should be corrected. The simplest method is, perhaps, to alter both the ordinate and the abscissa of any point on the original calibration curve of the in strument in the ratio where Po is the air density at the time of Pi calibration and is the new air density. In this way a point on the calibration curve for density is obtained, and by repeating this process for a number of points on the original curve the curve for density can be drawn.
As regards the effect of inclination to the wind direction, it has been shown that the axis of the vane circle may be out of align ment with the wind by an angle approaching 2o° before the read ings of the anemometer are in error by more than about two or three per cent. A fluctuating wind speed is another source of error, for the vane anemometer then tends to indicate more than the true mean speed, but errors will be negligible unless fluctua tions are very violent. If maximum and minimum air speeds are respectively 5o% above and below mean speed the error on the mean speed indicated by the instrument may be of the order of or i 5% high. It can be shown, however, that the error depends on the square of the amplitude of the fluctuations so it decreases rapidly as the wind becomes steadier. An amplitude of 5o% on either side of the mean represents a very unsteady air stream.
Electrical or hot wire anemometers have not been much used because of the elaborate apparatus and manipulation involved, but certain types have been developed for industrial application. They may yield excellent laboratory results, and are, in particular, well suited to the measurement of low air speeds. (See also AERIAL NAVIGATION.) (E. Ow.)