TIMBER BEAMS. 195 



small relatively to the depth it becomes of considerable 

 importance, and must be taken into account. In order to 

 avoid the complication due to this cause, it is best to 

 make the span as large as possible. For rectangular beams 

 the ratio of depth to span should not be less than one to 

 ten, and may with advantage be one to twenty. For girder 

 sections the span should be greater than this. 



107. Testing of Timber Beams. The most useful way of 

 testing specimens of timber is in the form of beams 

 supported at the ends, and loaded in the middle in a 

 similar way to those which have been mentioned. Timber 

 beam specimens may be of greatly varying sizes, from 

 large full-sized beams, on spans of 15 ft., to small speci- 

 mens 2 in. deep, 1 in. wide, and tested on a 36 in. span. 



b d? 

 For a perfectly homogeneous material ^ y may be taken 



W . L 



to be a sensibly constant quantity, and small beams yield 

 results just as reliable as those of larger ones. But timber 

 is not a homogeneous material, and, owing to the fact that 

 the grain is large relatively to the dimensions of a speci- 

 men, small pieces may yield results very different from large 

 pieces of the same kind of wood. For this reason it is not 

 advisable to employ pieces for testing purposes smaller 

 than those last mentioned. If a commercial test is to be 

 made on some one kind of wood a good many specimens 

 should be tested, and an average taken, because no two 

 pieces obtained from the same tree are quite alike. 



The results aimed at in a commercial test of this kind 

 are the central breaking load, and the deflection, either at 

 the maximum load or at a certain specified load. It is not 



I) d 2 

 unusual to calculate the value of - ' for the specimen in 



W . (j 



question, so as to provide a means of comparison with 

 beams of other dimensions, but of the same material. This 

 is also done in the case of some of the metals. For instance, 

 in the case of the cast-iron beam whose test has been given 

 above this value is as follows : 



b . d? 1-03 x (2) 2 1 



W.l == 3000 x 36 26,200' 



This serves as a convenient way of comparing specimens 

 of different dimensions. 



