27 



CORBELS. 



A corbel is deemed best for this case, as it has several advantages, 

 among which are the following: 



(1) It not only supports but unites the abutting stringers, forming 

 a portion of the bond between them. 



(2) It stiffens the joint, materially decreasing the strain in the 



stringers. 



(3) When employed single-span lengths of stringers may be used, 

 these being cheaper and more readily furnished than the double-span 

 length. The double-span length is of course better. 



(-1) Large beams, which stringers usually are, are particularly liable 

 to shear along their neutral axis. They will fail in this manner at less 

 than half the shearing strength per square inch, as indicated on a small 

 test. The reasons for this are: (1) That a large beam is apt to contain 

 a portion of the heart center of a log, which is likely to be ring shaken; 

 (2) with old trees— and the trees from which such sizes are cut are 

 usually old— the heart center is much below the average quality of the 

 cross section; (3) large pieces are particularly liable to check in season- 

 ing. Now, the holt through the corbel and stringer does excellent 

 service in increasing the resistance to failure by shearing. In fact, 

 even when corbels are not used, a bolt through the ends of the string- 

 ers would be a wise precaution. It would be necessary, however, to 

 tighten these occasionally until the timber had thoroughly seasoned. 



(5) Corbels in many cases furnish the only means of obtaining suffi- 

 cient bearing value for the stringers. 



Many companies will not use corbels, claiming — 



(1) That they increase the cost in labor, lumber, and iron. 



(2) That they increase the number of joints and hence the number 

 of places for the beginning of decay. 



(3) That they do not, after all, increase the bearing area of the 

 stringer, since, as the latter deflects, the whole load is brought upon the 

 ends of the corbel. 



With regard to the first item, by comparing the cost of designs 

 shown on figs. 2 and 3, we see that the corbel design costs in material 

 68 cents more per bent. The additional cost in labor would be very 

 nearly offset, probably, by the greater facility in handling smaller and 

 fewer pieces. Perhaps 75 cents would be a fair estimate of the extra 

 cost of corbel construction. It gives, however, much better structure, 

 and would undoubtedly secure enough additional length of life to more 

 than pay for the extra original cost. 



The second objection is of no more force than the assertion that a 

 chain containing eleven links is not so strong as one containing but ten. 



The third claim, in a corbel not exceeding 3 or 4 feet in length, is of 

 no moment. The deflection of a well-designed stringer at the end of 

 a 4-foot corbel will not exceed a quarter of an inch for the proof load. 



