The amount of air which may be made to pass through a ventilating flue of ordinary construction and of different heights is given in a table on page 45.

The available force for moving the air which is obtained by heating is very feeble, and quite likely to be overcome by the wind or external causes. Thus to produce the slight pressure equivalent to one tenth inch of water in a flue 50 feet in height would require a difference in temperature of 50 degrees. In a flue of the same height a difference of temperature of 150 degrees would produce the same velocity as that caused by a pressure of 0.5 inch of water. To produce the same velocity as that due to a pressure sufficient to balance 0.1 inch of water will require that the product of height of chimney and difference of temperature should be 1760.

It will in general be found that the heat used for producing velocity, when transformed into work in a steam-engine is considerably in excess of that required to produce draught by mechanical means. In a rough way, an increase in temperature of one degree increases the head producing the velocity only about one part in 500.

Ventilation by Mechanical Means is performed either by pressure or by suction. In the first case the air is increased in density and discharged by mechanical force into the flue, the flow being produced by an excess of pressure over that of the atmosphere, so that the air tends to move in the direction of least resistance, which is outward to the atmosphere. In the second case, pressure in the flue is less than that of the atmosphere, and the velocity is produced by the flowing in of the outside air. By both processes of mechanical ventilation the air is supposed to be moved without change in temperature, and the force for moving it must be sufficient to overcome effects of wind or change of temperature, otherwise the introduction of air will not be positive and certain. The velocity in feet per second for various differences of pressure is computed as explained in Article 32, and tables are given on pages 42 and 45 for use in computing the amount discharged per square foot of the area of the cross-section of the flue.

31. Measurements of the Velocity of Air.-The velocity of air or other gases is measured directly by an instrument

called an anemometer, or it is measured indirectly by difference of pressure. The anemometer which is ordinarily employed for this purpose consists of a series of flat vanes attached to an axis and a series of dials. The revolution of the axis causes motion of the hands in proportion to the velocity of the air. In the forms shown in Figs. 10 and 11 the dial mechanism can be started or stopped by a trip arranged conveniently to the operator. In some instances the dial mechanism is operated by an electric current, in which case

For measur

it can be located at a distance from the vanes. ing the velocity of the wind an anemometer, which consists of hemispherical cups mounted on a vertical axis, is much used. The anemometers are all calibrated by moving them in still air at a constant velocity and noting the readings of the dials. This is usually done by mounting the anemometer rigidly on a long horizontal arm which can be rotated about a vertical axis at a constant speed.

When the pressure is light it can be measured by using a U-tube partly filled with water. Such an instrument is shown

G

a

b c

E

FIG. 12.-U-SHAPED WATER
GAUGE.

D

in Fig. 12, attached to a flue. There being less than atmospheric pressure in the flue K, the water rises in the leg FE and sinks in the leg DE. The difference of level in the twolegs is ab, which is usually measured in inches. If the flue is under pressure the water will stand higher in the leg DE than in FE, but the method of use is essentially the same in all cases.

In case the pressure and velocity are great, considerable error will be made by using the open tube as above, and for such a case a Pitot's tube arranged as shown in Fig. 13 should be used.

This tube consists of two parts, one of which is straight and enters at right angles to the current dB; the other is curved so as to face the current at right angles, cA. These are connected to a U-shaped manometer containing water or some light liquid. The pressure in the two tubes will be the same except for the velocity of the current. This will tend to make the liquid stand higher in the arm fm than in the arm en. The difference in elevations of these two arms will be the velocity-head producing the flow. Call this difference in height h, and the ratio of specific gravity of the liquid in the tube and of the gas in the flue r; then will v√2ghr. That is, the velocity is equal to 8.03 multiplied by the square root of the difference in height multiplied by ratio of weight.

in case water is used in the manometer and the gas is air

at a temperature of 60 degrees, r will equal 813. Hence v will equal 228 √, in which is in feet, and will equal 65.7 h when his in inches of water. For any other temperature than 60 degrees this quantity must be multiplied by the square root of 460 + the temperature, and then divided by V520. Practically for air the velocity will equal 228 times the square root of the difference in the heights of the columns.

The velocity of air may also be computed by the heating. effects, provided the amount of heat is accurately measured

FIG. 13.-SKETCH OF PITOT'S TUBE FOR GREAT PRESSURES.

and the increase in temperature of the air be known. The specific heat of air is 0.238, hence the heat sufficient to warm one pound of water would heat (1/.238) = 4.2 pounds of air. This at 60 degrees would correspond to about 231 cubic feet. By consulting Table VIII the volume heated I degree by I heat-unit at any other temperature can be found.

The total number of cubic feet of air heated would be equal to the total number of heat-units absorbed divided by the number of degrees the air is heated, and this result multiplied by the volume of one pound divided by the specific

heat (the latter number can be taken directly from Table VIII). Having the total amount of air in a given time, the velocity can be obtained by dividing by the area of the passage.

NOTE. In the shape of a formula these results are as follows: Let Tequal temperature of discharged air, t that of entering air; H equal the total number of heat-units given off per unit of time; V equal the number of cubic feet of air heated 1 degree by 1 heat-unit (see Table VIII); A equal area of passage in square feet; v equal velocity for the same time that the total number of heat-units are taken. Then we shall have

32. The Flow of Air and Gases.-The flow of air obeys the same general laws as those which apply to liquids. The gases are, however, compressible, and the volume is affected very much by change of temperature, so that the actual results differ considerably from those obtained for liquids. These laws can only be expressed in mathematical formulæ, from which, however, practical tables are derived.

The flow of air from an orifice takes place under the same general conditions as those of liquids, and we have the general formula v = W2gh as applicable. In this case h is the head which is equal to the height of a column of air of sufficient weight to produce the pressure. Air under a barometric pressure of 30 inches and at 60 degrees in temperature is 813 times lighter than water. The pressure of air is usually measured by its capacity of balancing a column of water in a Ushaped tube (see Article 31), and this pressure is expressed in inches of water. One inch of water-pressure is equivalent to 65.7 feet of air at 60°, and increases part for each degree of increase in temperature. The above formula is only approximate, and does not account for the change in temper atures and of pressures due to expansion, although sufficiently accurate for the designing of ventilating apparatus. Prof. Unwin gives in the article "Hydromechanics," Encyc. Brit.,