FIG. 36.-BASKET-HANDLE AND U-SHAPED SECTIONS.

FIG. 37.-EGG-SHAPED AND RECTANGULAR SEWERS.

This illustrates the large amount of material used in placing an egg-shaped sewer on a flat foundation in unstable soil. In the rectangular sewer, the left side is designed for rock cut, the right for earth.

The above are a few of the forms which have been designed to meet special construction conditions. In the case of a combined sewer, the invert should be such that the dry-weather flow will be carried in a channel narrow enough to give a desirable depth of flow and velocity. This eliminates the flat bottom, although a bottom approximately flat but with a depressed channel for the dry-weather flow may be employed. In such case the channel should have sufficient depth so that the maximum amount of dry-weather flow will not overflow onto the benches. The most serious objection to this form from an operating stand

point is the probability that after a rain-storm, as the amount of combined sewage decreases to the dry-weather flow, suspended organic matters will be left stranded on the side benches. To prevent this, it is desirable to make these benches as steep as practicable. The dry-weather channel should be made as smooth and uniform in section and as straight in line as possible, an excellent plan being to line the bottom with split vitrified pipe embedded in the concrete, the walls of the channel being carried up from this channel pipe a distance equal to at least the radius of the pipe; a batter of as much as 45 degrees in these walls offering several advantages.

In the case of outfall sewers, intercepting sewers and others where there is at all times a considerable volume of flow, there is no serious objection to making the bottom of the sewer approximately flat, although such form gives a somewhat lower velocity than the circular.

Where the sewer is built on yielding soil, the bottom is generally given the form of a flat inverted arch to resist the upward thrust of the soil.

In all forms, angles should be avoided in the bottom or between the bottom and the side-walls, since angles greatly reduce velocity of flow and so encourage deposits; also because it is more difficult to remove deposits from angles than if the surfaces be joined by a cove or short-radius curve. Where the sewer is built of concrete, such curve also lessen the probability of cracking at the angles.

The inside surface of the sewer should be made as smooth as possible, to prevent the adhering thereto of suspended matters and also to afford as little retardation of velocity as possible on the flat grades. For this reason, forms for the inside of concrete sewers should be made of steel plates or, if of wood, of tongue-and-groove material closely driven together and greased. In addition to this, on the removal of the forms the interior of the sewer should be finished off by hand, removing all fins or other projections, filling the cavities with cement mortar, etc. In some cases, the walls have been rubbed down shortly after the removal of the forms to insure a perfectly smooth surface.

Where egg-shaped sewers are constructed, a smooth invert is sometimes obtained by embedding in the concrete, vitrified sewer-pipe split into thirds, this giving approximately the arc of the small invert-circle of egg-shaped sewers. There would

seem to be no advantage in this over monolithic concrete construction, however, with careful setting of invert forms and hand-finishing of inverts, combined with care in proportioning and mixing the concrete.

Where a large sewer changes direction, such change should be made by a curve and never by an angle, and the longer the radius of the curve the better. When two sewers join, one or both should be curved in the direction of flow of the other. In general, when a branch sewer joins a larger one, it is desirable to make the inner wall of the former tangent to the corresponding wall of the latter. If the branch is very much smaller than the sewer that it enters, however, the curve may be omitted and the branch make an angle of 45 degrees with the main sewer. Each junction of large sewers will generally require special careful designing. The sewer below a junction will ordinarily have an area approximately equal to the sum of the areas of all the sewers uniting at that point, and efforts should be made to have the sewage from all of these enter the sewer receiving their combined flow in a direction as nearly as possible parallel to the axis of such sewer. Also, it is desirable to so adjust the invert elevations and horizontal diameters as to have the surface of flow in all sewers at the junction point stand at the same elevation at all conditions of flow. Perfect attainment of this is generally impracticable, but the desirability of it should be borne in mind. Great improvement in this respect would have been possible in a great many sewer junctions that have been constructed.

Illustrations of the above points are shown herewith, accompanied by suggestions as to the reasons for the special form in each case.

The thickness that should be given the walls of a brick or concrete sewer should theoretically depend upon its shape, size, material, the pressure to be sustained contributed by or through the surrounding soil, etc. Brick sewers have frequently been

made one ring (4 inches) thick up to 30 inches diameter, two rings up to 60 inches, and three rings up to 120 inches. In the case of flat inverts on yielding soils, the invert should be designed to resist, as arch or beam, the vertical pressure exerted by the side-walls, unless these have bases sufficiently broad to prevent settlement independent of any support by the invert. The side-walls, or side arches, must be able to withstand the pressure of soil without; and if the soil can not carry the thrust of the arch without appreciable yielding, the side-walls should be able to carry this also.

Several formulas have been proposed for giving the thickness of arch of concrete sewers. Wm. B. Fuller's rule is: For crown,

d

12

+1 inch; for invert, 1 inch thicker; for haunches, 2 times

the crown; 4 inches being the minimum for the crown, 5 inches for invert and 6 inches for haunches (d being the diameter). C. D. Hill (chief engineer of sewer construction of Chicago) used the formula 0.28√R+0.1 foot, which gives approxi

For spans under 20 feet, the American Civil Engineers' Pocketbook gives, for thickness of arch at crown 0.04(6+S) for plain concrete, and 0.03(6+S) for reinforced concrete, S being the span of the arch. At the springing line, the thickness should be 50 per cent greater for circular, parabolic and catenary arches having a ratio of rise to span less than 1 : 4; and Ico per cent, if the ratio of rise to span exceeds this.

Buel & Hill, in "Reinforced Concrete," suggest a thickness at the crown of a concrete arch 0.0075 (S+10R), in which S is the span and R is the rise of the intrados.

Walter C. Parmley recommends a thickness of brick arch on a

horizontal line through the center of a sewer equal to

S

S

+2.572.

14

In general, reinforced concrete sewers, especially those approximately circular in section, have been made as thick as those not