126 Bulletin 86 



supporting strength of tile in ditches* ; the bedding of the tile in 

 the sand is not difficult or onerous, and this method should be pre- 

 ferred to the other methods in use, all of which depend upon doubt- 

 ful conversion factors. 



Downward pressure was provided by means of a lever. The ful- 

 crum was fastened to a tree stump, and the free end carried a 

 suspended box into which pails of sand could be poured. The ratio 

 of lever arms was one to five. Proper allowance was made for the 

 weight of the apparatus resting on the pipe. The time required, in- 

 cluding the measurements, was from 25 to 50 minutes per pipe. 



The internal diameters were measured vertically and horizon- 

 tally, and the wall thickness was measured on top, bottom, and each 

 side at the tongue end. Each pipe was weighed before testing. The 

 data of the tests in 1917 are given in Table XL 



The 12-inch, 16-inch, and 20-inch machine-made pipe, tested dry, 

 .were of almost identical strength. The breaking loads ranged from 

 4400 to 6300 pounds. The 14-inch pipe were weaker, the average 

 breaking load being 3440 pounds. This is due in part to the age 

 of the pipe, but an additional reason for the low strength is evi- 

 dent from an inspection of the wall thicknesses. To be consistent, 

 the 12-inch pipe might be one-eighth inch thinner or the 14-inch pipe 

 an eighth inch thicker. An accurate comparison can be had by com- 

 puting, for each group of pipe tested, the modulus of rupture, that 

 is, the maximum stress in pounds per square inch in the pipe shell 

 at the moment of failure. This comparison is made in Table Xllf. 

 Inspection of the last column shows the highest modulus of rupture 

 for the 12-inch pipe and the least for the 14-inch pipe. Apparently 

 there is considerable gain in strength after the first month. Exclud- 

 ing the 14-inch pipe, which was only one month old when tested, 

 the evidence is that the smaller the pipe the better and more densely 

 it was packed. This is perhaps characteristic of packer-head pipe 

 and indicates that the packer-head principle is better adapted to 

 small sizes of pipe than to large sizes. All the moduli are high as 

 compared with other tests on 1 :3 cement mortars. 



♦Report of Investigations on Drain Tile of Committee C-G of the American 

 Society for Testing Materials, published as bulletin of Iowa State College of Agri- 

 culture and Mechanic Arts. Vol. XII, No. 34. p. 102. 



tFor method of computation, see Ibid., page 87. 



