<k LE PONT VIERENDEEL." XXXIX. 



Such bridges are calculated on the hypothesis that the inter- 

 sections of the latticing, and the flange are articulated. Now, 

 whilst this is allowable in the case of a pin-connected bridge, it 

 is not only incorrect in the case of a rivetted structure, thereby 

 leading to an erroneous conclusion, but through it no account is 

 taken of the secondary stresses due to the bending which arises 

 from the rigid connections. 



A method of arriving at the value of these secondary stresses 

 has been devised by Mauderla in 1880*, who bases his system 

 on the hypothesis of the perfect fitting of the lattice bars with 

 the flanges, or, in other words, the invariability of the angles, 

 under different conditions of loading, of the lattice bars between 

 themselves. In order to reduce these secondary stresses to a 

 minimum, Continental engineers have advocated and designed 

 trellis girders with redundant members ; such have been erected 

 on the Rotterdam- Amsterdam Line, of 323 feet span. 



French Government engineers have investigated the question 

 during the years 1893-94, and as a result advocate the use of 

 multiple trellis, approaching as near as possible a plate girder. 



The Professor, however, states that these solutions are not 

 the proper ones for the purpose of getting over the difficulty due 

 to secondary stresses, and believes that the true solution will be 

 found, not in the complication, but the simplification of the 

 trellis, that is by doing away with all diagonals and reverting to 

 a type which he calls rectangular. The Professor states that in 

 a triangular trellis the diagonal is indispensable if the inter- 

 sections are articulated, but are redundant, and therefore useless 

 if they are rigid, 



The Professor claims three advantages for his type over all 

 other types, as follows : — 



1st. Theoretic advantage, in that all calculations can be 

 made without recourse to any hypothesis, and, further, that the 

 calculation is incomparably more simple than that based on the 

 theory of Manderla. 



*H. Manderla, Die Berechnung der Sekundarspannungen. Allgemeine Bauzeitung, 1880. 



