Зображення сторінки
PDF
ePub

the vibration of each wall would then be only fifteen-thousandths of a foot either way from the perpendicular, - a variation so small, and so gradually attained, that there is no danger in imposing it upon the side-walls by firmly fastening to them each shoe of the rafter. Expansion is also provided against by fastening down one shoe with wall-bolts, and allowing the other to slide to and fro on the wall-plate without rollers.

Fig. 13.

After the trusses are up, there are various ways of constructing the roof itself. If the roof is to be of slate, it is best to space the trusses about seven feet apart, and use light angle-irons for purlins, which are spaced from seven to fourteen inches apart, according to the size of the slate. On the iron purlins the slate may be laid directly, and held down by copper or lead nails clinched around the

Fig. 19.

angle-bar; or a netting of wire may be fastened to the purlins, and a layer of mortar spread on this, in which the slates are bedded. When greater intervals are used in spacing rafters, the purlins may be light beams fastened on top or against the sides of the principals

with brackets, allowance always being made for longitudinal ex pansion of the iron by changes of temperature. On these purlins are fastened wooden jack-rafters, carrying the sheathing-boards or laths, on which the metallic or slate covering is laid in the usual manner; or sheets of corrugated iron may be fastened from purlin to purlin, and the whole roof be entirely composed of iron.

When the rafters are spaced at such intervals as to cause too much deflexion in the purlins, they may be supported by a light beam placed midway between the rafters, and trussed transversely with posts and rods. These rods pass through the rafters, and have bevelled washers, screws, and nuts at each end for adjustment. By alternating the trusses on each side of the rafter, and slightly increasing the length of the purlins above them, leaving all others with a little play in the notches, sufficient provision will be made for any alteration of length in the roof, due to changes of temperature.

[ocr errors][merged small]

When wooden purlins are employed, they may be put between the rafters, and held in place by tie-rods on top, and fastened to the rafters by brackets; or hook-head spikes may be driven up into the purlin, the head of the spike hooking under the flange of the beam, spacing-pieces of wood being laid on the top of the beam from purlin to purlin. The sheathing-boards and covering are then nailed down on top of all in the usual manner.

CHAPTER XXVIII.

THEORY OF ROOF-TRUSSES.

In this chapter it is proposed to give practical methods for computing the weight of the roof with its load, and the proportion of the truss and its various parts.

The first step in all calculations for roofs is to find the exact load which will come upon each truss, and the load at the different joints. The load carried by one truss will be equal to the weight of a section of the roof of a width equal to the distance between the trusses, together with the weight of the greatest load of snow that is ever likely to come upon the roof. In warm climates, of course, the weight of snow need not be provided for.

It is a very common practice to assume the maximum weight of the roof and its load at from forty to sixty pounds per square foot of surface; but, while this may be sufficiently accurate for wooden roofs, it would hardly answer for iron roofs, where the cost of the iron makes it desirable to use as little material in the truss as will enable it to carry the roof with safety, and no more.

The weight

of the roof itself can be easily computed, and a sufficiently accurate allowance can be made for the weight of the truss; and, if the roof is to be in a climate where snow falls, a proper allowance must be made for that; and, lastly, the effect of the wind on the roof must also be taken into account.

Mr. Trautwine says, that within ordinary limits, for spans not exceeding about seventy-five feet, and with trusses seven feet apart, the total load per square foot, including the truss itself, purlins, etc., complete, may be safely taken as follows: Roof covered with corrugated iron, unboarded If plastered below the rafters

Roof covered with corrugated iron or boards
If plastered below the rafters.

Roof covered with slate, unboarded, as on laths
Roof covered with slate on boards 14 inches thick
Roof covered with slate, if plastered below the rafters.
Roof covered with shingles on laths

If plastered below the rafters, or below tie-beam
Roof covered with shingles on 7-inch board

8 pounds.

18

66

11

66

18

66

[ocr errors]

13

[ocr errors]

16

66

26

66

10

66

[ocr errors]
[blocks in formation]

For spans of from seventy-five to one hundred and fifty feet, it will suffice to add four pounds to each of these totals.

The weight of an ordinary lath-and-plaster ceiling is about ten pounds per square foot; and that of an ordinary floor of an inch and one-fourth boards, together with the usual three by twelve joist, fifteen inches apart from centre to centre, is from ten to twelve pounds per square foot. White-pine timber, if dry, may be considered to weigh about twenty-five pounds; Northern yellow pine, thirty-five pounds; and Southern yellow pine, forty-five pounds per cubic foot. Ordinary spruce may be considered to weigh twentysix pounds per cubic foot. Oak may be reckoned at from forty to fifty pounds; cast-iron, at four hundred and fifty pounds; and wrought-iron, at four hundred and eighty pounds per cubic foot.

For flat roofs, the weight per square foot of the various roofing materials on seven-eighths inch boards, not including the rafters or joist, may be taken as follows:

Roof covered with tar and gravel over 4 thicknesses of felt, 9 lbs.

[merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small]

From this data the weight of the roof itself may be easily computed, and we have then only the weight of the snow and effect of winds to allow for.

Snow. Any allowance for the weight of snow must depend upon the latitude. It may accumulate in considerable quantities, becoming saturated with water, and turning to ice. The weight of a cubic foot is very various. Freshly fallen snow may weigh from five to twelve pounds. Snow and hail, sleet or ice, may weigh from thirty to fifty pounds per cubic foot; but the quantity on a roof will usually be small. Snow saturated with water will usually slide off from roofs of ordinary pitch. An allowance of from twelve to fifteen pounds per square foot of roof will suffice for most latitudes.

Wooden trusses frequently support an attic-floor, and in such cases the weight of the floor and its greatest probable load should be considered as applied at the joints of the truss.

Wind Pressure. The load on the roof, thus far considered, is a steady dead load, which of course acts in a vertical direction. But roofs are frequently subjected to great pressure from the force of the wind; and, as this can act only on one side of the roof at a time, it is an unsymmetrical load, and moreover it does not act vertically. The pressure of the wind on an inclined surface is always normal to the surface, no matter what the direction of the

wind: hence the resultant of the wind pressure must act in a direction normal (at right angles) to the face of the roof. In this country the wind seldom blows with a pressure of more than forty pounds per square foot on a surface at right angles to the direction of the wind; and it is considered safe to use that number as the greatest wind pressure.1 But the pressure on the roof is generally much less than this, owing to the inclination of the roof. The following table gives the normal wind pressure per square foot on surfaces inclined at different angles to the horizon, for a horizontal wind pressure of forty pounds per square foot.

[merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][subsumed][subsumed][subsumed][subsumed][ocr errors][subsumed][merged small][subsumed][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small]

Until of late years it has been the general custom to add the wind pressure in with the weight of snow and roof; and, although this is evidently not the proper way to do, yet for wooden trusses it gives results which are perhaps sufficiently accurate for all practical purposes; and, if caution is taken to put in extra bracing wherever any four-sided figure occurs, this method will answer very well for wooden trusses. For iron trusses, however, the strains in the truss due to the vertical load on the truss, and those due to the wind pressure, should be computed separately, and then combined, to give the maximum strains in the various pieces of the truss. It should be borne in mind that a horizontal wind pressure of forty pounds per square foot is quite an uncommon occurrence, and, when it does occur, generally is of short duration; so that a truss which would not withstand this pressure, if applied for a long

1 At the observatory, Bidston, Liverpool, the following wind pressures per square foot have been registered. 1868, Feb. 1, 70 pounds; Feb. 22, 65 pounds; Dec. 27, 80 pounds. 1870, Sept. 10, 65 pounds; Oct. 13, 65 pounds. 1871, March 9, 90 pounds. 1875, Sept. 27, 70 pounds. 1877, Jan. 30, 63 pounds; Nov. 23, 63.5 pounds. - AMERICAN ARCHITECT, vol. xv. p. 237.

« НазадПродовжити »