THE FLOW OF WATER IN CONCRETE PIPE. 71 



DESCRIPTIONS OF PIPES. 



The following descriptions apply to experiments conducted by the 

 writer for the Bureau of Public Roads. The pipes were laid on a 

 hydraulic gradient and in no case were they flowing full of water. 

 For descriptions of experiments conducted under similar conditions 

 by other agencies see Appendix, page 77. 



No. 49, Experiment $-37. — 10-inch jointed concrete pipe chute- 

 drop on L-Y-7 line, Tieton project, United States Reclamation Service, 

 Washington. — Some of the topography of the Tieton project is such 

 that water for irrigation must be conveyed down very steep slopes 

 from one open channel to another. For this purpose chute drops of 

 both the open type and the pipe type are used. A series of experi- 

 ments was conducted on one of the pipes. 



Water from the upper canal is delivered through the bank into a 

 pool just above a 4-foot Cipolletti weir. After the water falls over 

 this measuring weir it drops down a 16-inch intake well about 2 feet 

 deep. At the bottom of this well the pipe chute, 10 inches in 

 diameter, leads down the hill; falling 87.6 feet in a developed distance 

 of 935 feet, 



The volume of water, Q, was determined by hook-gauge meas- 

 urements of head on the 4-foot weir, where the contraction condi- 

 tions were good and the velocity of approach reduced by a 

 brush screen. All of the items in Table 11 are based on the general 



hydraulic assumption that A = -&• As we have determined Q and V, 



we can solve for A as for a segment of 10-inch pipe. The resulting 

 retardation factors for various formulas are all right in so far as they 

 might be used in computing the velocity down a similar chute, but 

 may not be used in the computation of maximum capacity for the 

 following reason. All complete experiments upon friction losses in 

 chute drops of either the open-channel or pipe type have developed 

 the fact that the measured area of the wet cross section is much 



greater than -y- By a "complete experiment" is meant one in 



which Q,V, and the wet cross section, which we will call A' , were 

 measured independently. Such measurements disclose the fact that 

 Q = (A'-A")V where A" is an area made up of the aggregate 

 entrained air bubbles. As a concrete example take observation 6 

 (see Table 11, p. 68). The velocity, V, as measured was 13.59 feet 

 per second. The quantity, Q, as measured over the weir, was 4.29 



second-feet. What might be termed the net water area A=-$rWas 



0.316 square foot. The nominal area of a 10-inch pipe is 0.545 

 square foot, but as most small pipes run under size the actual area 

 was probably nearer 0.5 square foot. Thus, while by computation 

 but 54 per cent of the cross section of the pipe was filled, yet the 

 intake pool was full and no more water could be crowded into the 

 chute. In other words, the pipe was filled with a mixture of air and 

 water. At the outlet of pipe chutes running to capacity this fact is 

 manifest by periodic rushes of air into the outlet pool. 



It is not feasible and perhaps not possible to determine the mean 

 area of the water section down a pipe chute by actual measurements. 

 If such a thing were possible, then retardation factors could be com- 



