THE FLOW OF WATER IN CONCRETE PIPE. 13 



As made in local pipe yards along the Pacific coast, the smaller sizes 

 of pipe, say from 6 to 20 inches in diameter, appear to run under size. 

 The writer measured two diameters on many sections of pipe, taken 

 at random in various pipe yards. The results may be summed up 

 as follows : 



The average true area of 8-inch pipe was 4.4 per cent less than the nominal; of 

 10-inch pipe was 2.7 per cent less; of 12-inch pipe was 3.5 per cent less; of 14-inch 

 was 1.3 per cent less; of 18-inch was 1.4 percent less. Sections of greater diameter 

 than 20 inches appear to run quite true to nominal size. 



SCOPE OF THE EXPERIMENTS. 



The writer conducted 130 observations on 30 separate pipes, 29 of 

 which ranged from 8 inches to 63 J inches in diameter and one 120 

 inches in diameter. Seventeen pipes were of the " dry-mix," cement- 

 washed, jointed types; 5 were of the "wet-mix," oiled-form, uncoated 

 Jointed type; 3 were constructed in the same manner and then washed 

 with cement; 1 was of the wet-mix, monolithic, steel-form, coated 

 type; 3 were of the wet-mix, monolithic, wood-form, uncoated type; 

 1 of the same construction, coated. All but two of these pipes were 

 " running full ' ' ; that is, under pressure. Mean velocities ranged from 

 less than 1 foot per second to more than 9 feet per second. 



From other sources also listed in summary Tables 3, 4, and 11, and 

 briefly described in the Appendix commencing on page 77, descrip- 

 tions of experiments on pipe up to 18 feet in diameter are abstracted. 



EQUIPMENT AND METHODS EMPLOYED FOR COLLECTING AND INTER- 

 PRETING FIELD DATA. 



With the following exceptions, the equipment and methods used 

 were the same as those employed in the experiments on wood-stave 

 pipe and which were described in Bulletin 376. of the department. 

 For the sake of brevity these descriptions will not be repeated here. 



Color injector. — The only practicable method of measuring, the 

 velocity of water in some of the pipes tested was by timing the passage 

 of some color or chemical. For comparatively small volumes of clear 

 water a saturate solution (from 5 to 8 per cent in air temperatures) of 

 potassium permanginate was used. This color is rather uncertain, 

 being decomposed by some waters and neutralized by reddish or 

 muddy waters. Where certainty of results is desired, almost regard- 

 less of the water, fluorescein is the best chemical and Congo red the 

 second choice, in the knowledge of the writer. At the present time, 

 however, either is practically impossible to obtain. 



It is desirable to start the color at a gulp, whether in water under 

 pressure or in a pipe but partly filled. The "color gun" used in the 

 experiments upon wood-stave pipe was improved somewhat. As 

 shown in figure 3, S is a reservoir into which the solution is poured, 

 through the coupling W. The latter is then closed and the air reser- 



