table gives the depth of socket and the thickness of joint for the

two classes, the relative appearance being shown in Fig. 6.

2

121210

3

Experiments have been made on the relative value of the two kinds of sockets, and while it has been found in laboratory experiments that on account of the porosity of the cement filler, a wellmade joint in a wide socket allows a greater leakage than in a

standard socket, yet practically the increased space for the joint filler makes imperfect joints less likely, and really makes a tighter line. The deep-and-wide sockets should, therefore, always be used wherever the sewer is laid below ground water, and where consequently there is danger of leakage into the pipe.

Another recent improvement in the socket is the introduction of corrugations on the inside of the bell and on the outside of the spigot, by which the cement is held firmly and cannot be driven out by settlement or pressure. These corrugations are not added by all firms, but they are easily scratched on the pipe and should always be called for in specifications.

CHAPTER II.

SEWER PIPE, Continued.

Up to the year 1890, no comprehensive experiments on the strength of pipe had been carried out, and no systematic attempts to discover under what conditions' sewer pipe could be safely trusted to carry its given load had been made. The few tests recorded before that time are isolated experiments by engineers made in the course of their regular work. In 1859, Mr. Adams, City Engineer of Brooklyn, made some crushing tests of the material used in pipes. He prepared some two-inch cubes, and obtained a pressure of 50,000 pounds, the capacity of the machine, or 12,500 pounds per square inch, without crushing the material. He also applied pressure along the top of some domestic and imported pipes, and found that they broke as follows:

In 1878, Mr. J. Herbert Shedd,2 City Engineer of Providence, made some tests on the strength of standard sewer pipes, halfbedded in sand, with the following results in pounds per linear foot of pipe.

In 1890, Mr. Malverd A. Howe,' of the Rose Polytechnic Institute, undertook to make systematic tests that would be comprehensive, so far as American pipes were concerned, and for this purpose he obtained in the open market specimens of pipe from the different factories between Wilmington, Del., and St. Louis, Mo., fifteen different firms being represented. The pipes were subjected to five different kinds of tests, viz., hydrostatic, drop, concentrated load, uniform load, and joints.

The hydrostatic tests were made to find out the strength of the pipe against internal pressure, the ends of single lengths of pipe being closed and water pumped in until the pipe broke. The average tensile strength of the material for the different sizes was as follows:

The number of specimens tested for all sizes up to 18 inches was 25, and but two above 18 inches. From these results the experimenter concluded that the average tensile strength of the material composing American vitrified sewer pipe was at least 600 pounds per square inch.

Most of the pipes broke at an internal pressure of about 100 pounds per square inch, and the following table shows the computed thickness of the various sizes, assuming an internal pressure of 100 pounds per square inch, with a tensile strength of the material of 600 pounds, as compared with the thicknesses now made commercially.

The table shows that the thickness of sewer pipe is such that pipes will stand an internal pressure of 100 pounds and even more

with the smaller sizes before bursting, or a safe pressure of 33 pounds with a factor of safety of three.

The drop test was made to determine the resistance of the pipe to percussive action, such as a blow from a wagon wheel. It was made by supporting the pipe on two pieces of wood 2 inches wide, 16 inches apart, so arranged that a falling weight would strike the pipe near its center, midway between the supports. The weight was a box full of iron, weighing 18 pounds. A rounded strip of wood on the bottom was the striking part. The length of the drop was adjustable, but was 12 inches for the first five blows. If the pipe was not then broken, the length of drop was made 18 inches, then 24 inches, with 30 inches as a maximum. Twenty lengths were broken at the first blow, and most of the pipes were broken in four to ten blows. Mr. Howe's conclusion was that sewer pipe as made is strong enough to sustain ordinary blows, but it is evident that where successive blows may be expected, ample covering of earth or similar material should be provided to distribute the shock.

The concentrated load test was made by supporting the pipe, as just described, and then slowly applying the load through the medium of an hydraulic piston, acting against a small block of wood at the middle of the top of the pipe. Forty-two pipes of various sizes were broken, and while the smaller sizes withstood much more than 2000 pounds, it seemed a safe conclusion that the average pipe would stand at least that amount concentrated at the center with the supports 16 inches apart.

The uniform load test was made by bedding the pipe in sand in a strong box and applying pressure through a sand cover. Most of the pipe failed by splitting longitudinally at the top, bottom, and sides, and after splitting and taking their new bearings, were able to carry much heavier loads. The breaking loads, however, were taken when the pipe cracked. The small sizes. sustained a load of about 8000 pounds per linear foot of pipe, and the larger sizes a little over 2000 pounds, the conclusion being that all sizes of pipe will stand a load of 2000 pounds per linear foot before breaking.

In 1897, Mr. Barbour, then City Engineer of Brockton, Mass., made some experiments' on the strength of pipe by covering it with about a foot of earth and applying the pressures by means of an hydraulic piston pressing down upon the earth cover. He found that the breaking load per linear foot averaged about 2800 pounds for standard pipe, and about 4200 pounds for double strength pipe. He also studied the relation between the strength and thickness and concluded that the strength varied inversely as the diameter and directly as a function of the thickness, the relation being approximately expressed by the equation

P

=

C

11.65
d

where P is the pressure in pounds per linear foot,

t is the thickness in inches, and C a constant equal to 33,000. The table shows the relation obtained experimentally and by the formula, and their close agreement.

Mr. Barbour concludes from his experiments that manufacturers should be able to produce, and that engineers should demand, pipe which would have a breaking load of 3000 pounds per linear foot for standard pipe, and of 4500 pounds for double strength pipe, the thickness being so varied according to his formula or otherwise, that this strength should be obtained in all sizes. The thickness thus required is given below and may be compared with the thicknesses given in the table on page II: