THE FLOW OF WATEK IN CONCRETE PIPE. 75 



must be done from the inside alone under the shadows and imperfec- 

 tions of artificial lighting. 



For all shapes, circular or otherwise, the invert is usually laid first. 

 Forms for the remainder of the section are then placed on the invert 

 and the section completed. Sometimes the first pouring extends just 

 across the bottom and sometimes is carried up to the spring line. 

 This depends on shape and size of the section. Smooth joints 

 between the pourings and between the ends of adjoining set-ups are 

 difficult to attain. Unless very heavily braced, the pressure of the 

 head of wet concrete will spring the forms and result in offsets. On 

 the Catskill aqueduct this amounted to as much as 0.15 foot. 1 The 

 writer has seen cases where contractions of from 2 to 4 inches were 

 made in sections 5 feet or less in diameter. Other factors that influ- 

 ence this movement are weather, length of sections, rapidity of filling 

 the forms, and the type of aqueduct. 1 



Before the advent of steel forms, engineers anticipated a rough 

 interior when the forms were stripped and relied on a plaster coat to 

 fill interstices, estimates of cost being based on that of a plaster coat 

 on a certain sized conduit equated against that of an unplastered 

 surface on a larger conduit. 2 The present practice in general is to 

 use smooth steel forms; clean them thoroughly after each casting, 

 then dry and oil or grease them before using again. If close attention 

 is given to spading the mixture into close contact with the forms, then 

 a surface is secured that is smoother than will be attained by adding a 

 plaster coat, local roughness alone being improved by pointing up 

 and chipping down. The crown in particular is liable to be rough. 1 



Experience indicates that a given surface is more assuredly repeated 

 in a pipe of precast sections than of monolithic construction. Thus a 

 smaller factor of safety in the way of capacity overload may be assumed 

 for precast pipe. The experiments indicate that some of the smooth- 

 est pipe is of careful monolithic construction ; likewise some of the 

 roughest. Thus, while a very smooth surface may be attained, it 

 can not be anticipated with the assuredness possible where units 

 are precast under the best practice and subject to the conditions 

 attainable in a pipe yard. 



For all types of pipe, competent engineering supervision and 

 inspection are necessary if the best results are to be obtained. 

 Velocity and entry heads should be provided at the intake. Align- 

 ment and grades should be true. Changes in either should be made 

 with as gentle curves as are feasible. Unless it is excluded above 

 the intakes, means should be provided for the removal of debris at 

 the foot of steep upward slopes. Means for the removal of air, not 

 only at the summits but also near the intake (unless deeply sub- 

 merged) should be provided. This is true of small orchard lines as 

 well as large siphons. Trash racks, originally placed at the intakes 

 of many siphon pipes, have since been removed in many cases 

 because the danger from washout, due to the accumulation of trash, 

 outweighed the good accomplished. 



■ For all practical purposes it does not appear necessary to limit the 

 higher velocities. The velocity should be at least sufficient to 

 prevent the deposition of silt — probably not lower than 3 feet per 

 second. 



1 Water Works Handbook, by Flinn, Weston and Bogert, New York, 1916, p. 276. 



2 Jour. West. Soc. Engrs., Vol. XVI, No. 8, 1911, p. 696. Discussion of C. C Saner. 



