Document YNbJnORm8nrJk4YvDmyb8XmD

American Society of Heating and Ventilating Engineers Guide, 1934 jo too iso . too tso emit. hue Radius m Pulccht or Pipe O/amitcil Fig. 1. Curve Showing Loss of Pressure in Round Elbows 0.5 times the velocity head. The pressure loss in elbows must also be allowed for in the design. It is customary to express dynamic losses in terms of the percentage of the velocity head; in other words, the per centage of that pressure corresponding to the average velocity in the duct which is expressed in terms of inches of water gage. Figs. 1 and 2 show the effect of changing the radius of elbows of square and rectangular section. These charts are based on tests of pipe elbows of ordinary good sheet metal construction. For example, a five-piece round pipe elbow having a centerline radius of one diameter has a loss of about 25 per cent of the velocity head. At a velocity of 2000 fpm the corresponding head 256 Chapter 19--Air Duct Design is 0.25 in. water gage, and at this velocity the elbow just referred to would cause a pressure drop of 0.063 in. water gage. Experience has shown that good results may be obtained when the radius to the center of the elbow is V/t times the pipe diameter. The pressure drop will then be approxi mately 17 per cent of the velocity head for round ducts, and 9 per cent for square ducts. Very little advantage is gained in making elbows with a radius of more than two diameters. Friction Losses Friction losses vary directly as the length of the duct, directly as the square of the velocity, and inversely as the diameter. Since length is a fixed quantity for any system, the factors subject to modification are the area and the velocity, which determine the relation between the first cost of the duct system and the cost of the power for overcoming friction. The friction between the moving air and pipe surface causes a loss of head which is numerically equal to the pressure required to maintain a given velocity, and is expressed in the following modification of Fanning's formula: For round pipe and standard air (70 F and 9.92 in. barometer) , _ , L , L ( V \2 hh f D "v CD \ 4005 ) (3) For rectangular ducts (4) 'where hh -- loss of head, inches of water. (V \2 4005 / = velocity head, inches of water. V -- velocity of air, feet per minute. L = length of pipe. D -- diameter of pipe. o, b = sides of rectangular duct. . / = coefficient of friction. I f all in feet J C =-i = length of pipe in diameters for one head loss. For all practical purposes C varies only with the nature of the pipe surface: C = 60 for perfectly smooth pipe; = 55 for pipe as used in planning mill exhaust systems; = 50 for heating and ventilating ducts; = 45 for smooth and 40 for rough conduits of tile, brick or concrete. However, Fritzche states (and numerous tests check very closely) that / varies inversely as the 2/7 power of the pipe diameter, and inversely as the 1/7 power of the velocity, or inversely as the 1/7 power of capacity, which is the same thing. Thus Formula 3 may be revised as follows, based upon a loss of one velocity head (at 2000 fpm) in a length equal to 50 diameters of 24 in. galvanized swedged pipe: L / V \ 13/7 Al = 1.1 CD"7 \4005 / (5) 257