As already said, the shearing strength of iron is equal to about 4/5 of its tensile strength. In order to cover all the changing values of unit-stress, and thus to avoid tedious work of calculation, we may do well to fix the unit stress for shear at about the lowest for ordinary rivet iron. Taking, then, the unit tensile stress at 9,000 lbs. (for c in formula 14), we obtain the allowable shearing stress of 9,000Ã--4/5=7,200 lbs.

We will assume that the shearing stress is uniformly distributed over the cross-section of the rivet. The strength of a rivet ⅞ inch in diameter for single shear will then be (⅞)2 Ã-- ¬ Ã-- 3.1416 Ã-- 7,200=4,300 lbs. And for double shear=8,600 lbs.

For a riveted joint auch as is shown in section by Fig. (19) the amount of stress on the rivet must neither exceed the double shear strength on the one hand, nor, be so great as to crush the rivet hole or the rivet itself on the other. The latter is usually the first to fail. It is a well-established fact that the allowable compressive stress per square inch on the projected area (diameter of hole X thickness of the plate) of the riret hole can be as much as and even more* than twice the shearing stress allowed, or 7,200Ã--2 = 14,400 lbs.

Fig. 19

If b - ⅜" we have for ⅞" diameter rivets the bearing value of ⅞ Ã-- ⅜ Ã-- 14,400=4,700 lbs.

Since we have no plate or angle in the girder whose thickness is less than ⅜ inch, and as the single-shear value of the rivet is less than its bearing value on ⅜" plate, we need not take bearing value into consideration, except when the rivets are double-shear.

Rivet holes are made either by drilling or punching. The former does not weaken the plate as much as the latter does, but drilled holes produce a cutting action on rivets unless their sharp, square edges are slightly rounded. It is, however, useless to enter here into any discussion on the relative merits of these two processes, as, in the present state of bridge building in America, viz., of competition in designs and prices, no bridge builder will go into the extra work of drilling and rounding rivet holes, when in a few minutes a longest plate or angle that comes out of a rolling mill can be punched from one end to the other, with all the accuracv and correctness of templet works. For this, however, a certain relation must always exist between the thickest plate used and the diameter of rivets connecting them If the plate be too thick, punch will be crushed before the plate is punched.

* Gerber has shown by experiment that this unit stress may be twice the unit tensile stress, or, in this case, equal to 9,000Ã--2 = 18,000 lbs.

If d=diameter of rivet hole to be made in plate of the thickness t, and c=crushing strength of the punch, while s=shearing strength of the plate; then the crushing strength of the punch ã/4 d2c must be equal to or greater than the resistance to shearing of the plate ndts.

Putting ã/4 d2c=ãdts d=4t 8/c.

As c is not much above 4s it is not advisable to make t more than d.

For steel plates t should, of course, be less than d of parts. In practice, to suit the general proportion of parts, the following diameters of rivets are to be recommended :

For spans up to 50 ft., ¾ in. diameter, and over 50 ft., ⅞ in.

In order to facilitate shop works it is quite essential not to use rivets of different diameters in a single piece of work. The diameter of rivet holes should not be more than 1/16 inch greater than that of the rivet. Every rivet should fill up the hole, and in case any signs of looseness are detected, should be cut and riveted anew.

Rivet Spacing

In order that a piece be punched with a machine it is evident that rivet holes should be spaced in straight lines, in directions both of width and length. With most machines it is necessary that no rivet spacing in a row should contain fractions less than ¬ inch.

Rivets in a row are spaced rarely closer than three times the diameter of rivets. Thus with ⅞" rivets, we take 2¾" as the minimum distance allowable between the centers of rivets in a row.

The maximum distance allowable between the rivets is determined from the consideration that, in the part of the girder subjected to compression, the thinnest plate composing that part shall be able to sustain the maximum stress, acting on its section without buckling between the rivets, by which they are held fast to other plates and angles. Generally, if spacing does not exceed twelve times the thickness of thinnest outside plate on the flange, we are on the safe side. Thus, in our girder, the least thickness of the plate being ⅜", we have for the maximum distance, center to center of rivets, 12Ã--⅜ = 4« inches.