to Transverse Strain in Beai 



423 



would appear from this that the maximum tensile stress might 

 have been taken to be somewhat less. But it must be remem- 

 bered that for convenience of calculation the stress has been 

 assumed to vary with the strain until the former has reached 

 its maximum amount, whereas in fact the overstrain com- 

 mences considerably sooner. The result is that the calculated 

 deflexion must be somewhat deficient, especially for the loads 

 about one half the breaking weight. Making allowance for 

 this circumstance, the calculations accord very fairly with the 

 facts as observed. 



Mr. Barlow relied much for the confirmation of his theory 

 upon the fact that the excess of strength exhibited by solid 

 beams does not appear to the same extent in flanged girders, 

 or in beams of a form which he had made for the purpose of 

 experiment in which an upper and lower flange were connected 

 only at intervals, the effective section being therefore that 

 shown in fig. 12, the space acdb being vacant 

 between the solid portions A a b B and cCD d. 

 But this will evidently be the same on the 

 theory of overstrain here proposed. 



In such a beam before the point of rupture 

 is reached, unless the intermediate space is a 

 very small fraction of the total depth, the whole 

 of the lower part of the section will be included 

 in the area of maximum stress, and if d x —md 

 be the depth of the upper and lower solid 

 portions, and the other terms and dimensions 



Tip a. 



as in the former case, then as before I K 



bd-h 



B K 



and (fig. 13) 



and the area of compressive stress Iz&K being of necessity 

 equal to the area of tensile stress c G D d. 



