ON BRIDGES — BEAMS. 



143 



FOURTH LECTURE. 



BRIDGES AND BEAMS 



[Continued.] 



The forms of triangular framing that we have noticed are not suited 

 to all cases, and we return to the double-flanged beam, and consider 

 its application to long spans. 



There are certain limits which cannot be passed in making beams 

 in a single piece, and recourse must be had to some arrangement of 

 connected pieces, which will be economical and effective. If we use 

 boiler plate we may make a composite beam of the same 

 form as the simple ones already described, as in Fig. 30, 

 the web being still a thin flat plate, and the flanges being 

 formed by riveting angle irons to it. In cast iron this 

 would be hardly practicable, owing to the difficulty of cast- 

 ing a thin plate of any great size. In wood it would be 

 entirely impracticable with any regard to economy of mate- 

 rial. 



As stated before, the web may be separated into two plates, 

 and the flanges made cellular ; but we may go further, and, retaining 

 the flanges, connect them by an open web, in which the material shall 

 be so disposed as to resist strains under the best possible conditions. 



In a beam thus made, we have a top and bottom chord or flange, 

 connected by pieces of timber reaching from one to the other. If 

 these pieces or posts are disposed, as in Fig. 31, they will not serve to 



Fig. 31. 



Fig. 32. 



connect the chords properly, since a weight applied will cause the 

 structure to deflect, as in Fig. 32, the posts merely transmitting a 

 portion of the strain to the lower chord, and the whole system having 

 no more strength than it would have possessed had the posts been 

 omitted, and the beam made of depth equal to the sum of that of the 

 two chords, while we desire to take advantage of the distance between 

 the chords to give greatly increased stiffness and strength. 



The shape of the spaces or bays is evidently altered by the deflec- 

 tion in Fig. 32 from rectangles, as in Fig. 31, to rhomboids, the two 

 diagonals of which are not equal. Now the rectangle abed cannot 

 change into the rhomboid a' b' d d f , without c b becoming shorter, and 

 a d longer. If, therefore, we can prevent sucli change of length, we 

 can preserve the shape of the figure, and prevent the sinking of the 



