classification of end posts and top chords of truss bridges is frequently low on account of the eccentricity of the member with respect to the location of the pin.

The pins of old truss bridges frequently show a startlingly low classification where computations are made in accordance with the usual methods; hence it is necessary to take advantage of certain conditions which are more favorable than the usual assumption, in order to help out the classification. Where eye-bar members consisting of more than two pairs of eye-bars meet on a pin, a slight redistribution of stress in the several eye-bars will frequently increase the classification of the pin; and this is justifiable as being in line with the way the structure actually works. Where certain members have wide bearing surfaces on the pins, the center of pressure can be taken near one edge of the bearing surface, thus increasing the classification of the pin and, at the same time, approximating more nearly to the actual behavior of the detail. It is also permissible to use higher unit stress for figuring pins than for the other members of the structure. The following illustrates what might be considered permissible, providing there is assurance that the material is of good quality and that the computations take account of all the forces acting:

It is to be noted that, in bridges built in the late 80's and early 90's, hard grades of steel were frequently used for the pins.

In timber-trestle bridges, the stringers in bending usually show low classification. On account of there being three or more sticks acting together, it is permissible to use a higher unit stress for trestle stringers than for a single stick, as the average strength for the several pieces exceeds that of the poorest one. On account of the exposure to the weather and the deterioration which gradually takes place, the allowed unit stress in timber stringers should be reduced as the age of the bridge increases. Where timber bridges are thoroughly inspected and defective material is promptly replaced, and where they are subject to the same general consideration as given above for metal bridges, the following unit stresses might be taken as a safe practice for maximum fiber stress in stringer bridges without allowance for impact:

For stringer bridges six years old, 2,000 lbs. per sq. in., and reduced about 100 lbs. per sq. in. for each year following.

The above figures are based on Douglas Fir or dense yellow pine and for climatic conditions prevailing in the North Central States. In more arid regions where longer life of timber may be expected, the reduction in stress for age need not be so rapid. On account of the comparatively short life of timber bridges and the ease with which they can be renewed, there is not generally the same urgency in establishing maximum-safe-stress limits as in the case of the more permanent metal bridges. In timber truss-bridges the lowest classification usually occurs in the floor beams, truss rods, and diagonal braces.

It has been found that metal bridges suffer frequently from corrosion in the top flanges of stringers and floor beams, on account of the action of brine drippings from refrigerator cars.

In bridges where the ties are supported on shelf-angles riveted to the webs of the girders, the shelf-angles frequently show considerable corrosion and tend to break in the root of the angle.

In pin-connected trusses, excessive wear sometimes takes place in the pin bearings, particularly in draw bridges.