4 BULLETIN" 854, U. S. DEPARTMENT OF AGRICULTURE. 



For comparison, 69 tests were made on 10 and 12 inch tile, so laid 

 as closely to approximate poorly laid drains as found in the field, 

 to show the results of using unskilled workmen in laying drains 

 without proper supervision. Nine experiments were made upon the 

 loss of head in catch-basins, using 8-inch clay tile. Grades of 0.20> 

 0.75, and 1.50 per cent were tested, with drops in the catch-basin of 

 0.10, 0.20, and 0.30 foot for each grade. 



The investigations were made at Arlington, Va., in 1916 and 1917. 

 The experimental plant was designed and constructed by S. W. 

 Frescoln, drainage engineer, and the tests were made by D. L. 

 Yarnell, senior drainage engineer, under the direction of S. H. 

 McCrory, chief of drainage investigations. S. M. Woodward acted 

 as consulting engineer for the investigation, making suggestions in 

 the conduct of the experiments and collaborating in the preparation 

 of the data and report. 



This report upon the investigation of flow in drain tile includes a 

 detailed description of the equipment and methods used and the 

 tabulated data from the experimental work. The results deduced 

 from the data are shown graphically, the method of developing the 

 curves being explained. The formulae now in general use for com- 

 puting velocity in tile are discussed and comparison is made with the 

 new formula presented . A diagram is given showing discharge capaci- 

 ties based upon this formula, covering sizes from 4 to 48 inch tile, 

 and grades from 0.04 to 3.00 per cent. 



So far as the writers have been able to learn, only one other similar 

 investigation of this subject has ever been made. This was by 

 Messrs. J. F. Rightmire and M. E. Chappel and was quite limited in 

 extent (see Vol. IV, No. 4, Bulletin of the Iowa State College Engineer- 

 ing Experimental Station). 



CONCLUSIONS. 



The following general conclusions have been drawn after a detailed 

 study of all of the experimental data : 



(1) That the value of the coefficient of roughness, n, in the Kutter 

 formula, as obtained by experiments in a drain or pipe at any depth 

 of flow less than full, does not necessarily apply to that drain or pipe 

 when flowing full. 



(2) That the exponent of the slope, s, is practically 0.5. In other 

 words, the loss of head is in proportion to the 2.0 power of the velocity 

 and not the 1.8 power, as given by many authorities. 



(3) That the exponent of the mean hydraulic radius, R, is 2/3. 



(4) The Chezy formula gives the same velocity of flow in a pipe 

 flowing one-half full as in one flowing full, with the grade constant. 

 The experimental data obtained seem to disprove this commonly 

 accepted theory. 



