804 PROF. A. H. GIBSON ON THE LOSS OF ENERGY AT OBLIQUE IMPACT 
formed being comparable as regards smoothness with the rest of the pipe. Under these 
circumstances the losses were considerably reduced. The plots of the logarithms of 
the loss on a base of logarithms of 6 (B B’, fig. 2) now lie on a straight line repre- 
senting the formula— 
loss= F ee feet, 
29 
where F = :00006760°™. 
The mean experimental values of F are given in the following table against those 
calculated from this formula :— 
6 BF 60° 45° 30° oy, ‘5S 
{ calculated 1:20 “495 264 +109 0294] 0023 
| eeeamenen 1-20 492 263 Ltt 0240 , — 
Werspacu,* from experiments on a series of elbows of circular, cross-section 
1-2 inches in diameter, obtained values of F corresponding to 
F="9457 sin? § 42-047 sint § | 
This gives the following values—_ 
6 90° 60° 45° 30° 15°: 
F ‘987 | +365 183 0728 ‘0222 
It will be noted that these are considerably less than the author’s values. The 
difference may be due to some extent to the difference in the shapes of the passages, 
although a few more recent experiments by BricnTmorEt and by Datey { on elbows and 
tees of circular cross-section give results in close agreement with those of the author. 
Thus Bricurmore found F equal to 1°19 for right-angled elbows in 3-inch and in 4-inch 
pipes, while Datry, using a 4-inch right-angled tee filled in to make a square elbow, 
found F equal to 1:10, and equal to 1°50 when the idle branch formed a dead end 
full of water. 
4. Loss at Impact or Two ConFrinep SrrREaAMS. 
In determining the loss due to the impact of two streams, experiments were carried 
out on each pipe (@) with v, constant, v, varying; (>) with v, constant, and v, varying. 
The loss per lb. of the impinging stream (2) was then calculated. An examination of 
the results shows that on plotting the loss with v, constant on a base of v.2+2g the 
* Mechanics of Machinery. + Proc. Inst. Civil Engineers, vol. elxix., 1906-7, p. 326. 
{ Cornell Civil Engineer, Dec, 1911, p. 107. 
