644 PROCEEDIXGS OF SECTION H. 



It is clear that, if diagonals be introduced at the centre panel in 

 Figs. 2 and 2a, the joints become very much more difficult to make, 

 especially that at the top of the post, and the beam becomes con- 

 siderably cut into and weakened. 



In trussed beams, then, so far as the timber work is concerned, 

 the bridge carpenters' work is of a very simple kind. The steel or 

 iron work would offer little difficulty to a blacksmith. If the rods 

 would be unwieldy for transporting one length, pin-joints could be 

 introduced at one or more points, but these add to the first cost. 



Other Imperfectly-braced Structures. 



The reversed trussed beam becomes an unbraced Queen-post 

 truss, the stresses in which can be calculated in a way quite similar 

 to that explained in the foregoing pages. The tension member is 

 here horizontal, and it is relied on to resist the bending action of 

 the loads. 



In the first place, it may be laid down that no tension member 

 should ever be made of timber, because no reliance can be placed 

 upon its resistance to shearing along the grain, in consequence of 

 the tendency of timber to open in longitudinal cracks. No ordinaiy 

 factor of safety will allow for this. But, on the other hand, if we 

 make the tension member of s);eel or iron, its section will not be a 

 suitable one for a beam to resist bending. 



Stringers, placed on cross-beams, may be made strong enough 

 to act as beams; but, in order that they may be most effective, they 

 must be anchored down at the ends. In fact, to make the stringers 

 behave, in relation to the Queen-post truss, in the same way as the 

 beam in the structure called the " ti-ussed beam," we should introduce 

 a stout cross-beam under the feet of the batter braces at each end 

 of the bridge, and we should securely bolt the stringers down to these. 



Or we may make the bottom member of the Queen-post truss 

 composite in character, the tension being taken by nietal and the 

 bending by the beam, the two members lying side by side, or the 

 steel on top of the beam. 



In all that has been written, it has been taken for granted that 

 n') initial stress has been introduced, such, for instance, as that 

 which would be caused by tightening up the nuts on the tension rods, 

 or such as would be caused by leaving the nuts on the rods slack, 

 thus producing initial bending stresses in the beain. These initial 

 stresses would, of course, modify the results, just as in the case of a 

 braced girder, having pin-joints, if the members are not jjerfectly 

 made, some may be too long, requiring to be compressed into their 

 places, and others may be too short, so requiring to be stretched, 

 thus modifying the values of the stresses calculated in the ordinary' 

 way. 



Should there be initial stress, it must be ascertained, or esti- 

 mated as accurately as possible, and its amount added to, or sub- 

 tracted from, those calculated, as may be necessary. 



The writer has to apologise for Ijringing a partially completed 

 paper before the meeting. It was with diffic\ilty that time was 

 found for carrying the investigation so far even as it has been 

 va'-ried. He felt the need of some simple treatment of tl:is subject. 



