i8o 



TOWERS AND TANKS FOR WATER-WORKS. 



must be designed in accordance with the results of experience. 

 In the case of the Fairhaven tower experience to point out features 

 requiring special attention in the design was lacking. Again, 

 very serious changes (see Engineering News, Nov. 21, 1901) 

 from the original design were made in the construction of the 

 tower, and I consider that these added greatly to the danger 

 of failure. The engineer who designed the structure cannot be 

 held responsible for its failure, in my opinion. 



" In Fig. 49 I have reproduced your sketch (with some additions) 

 showing the methods of supporting the tank at the tops of the 



posts. Owing to the posts having a 

 batter, the post thrust against the 

 bottom of the circular girder will have 

 an inward horizontal radial com- 

 ponent. From the original descrip- 

 tion of the tower published in Engi- 

 neering News of Sept. 5, 1895, I 

 judge that the batter was about i to 8. 

 With the tank full the horizontal 

 radial component would, therefore, 

 be about 36,000 Ibs., as indicated in 

 Fig. 49. No special provision appear;; 

 to have been made in the design to 

 take care of these radial pressures, 

 except as the lower flange of the 

 circular girder may be capable of 



// withstanding them. In all designs 



/ for water-towers these forces should be 



provided for, and usually are provided 

 for by the use of a circular girder with 



its web horizontal. The company which builds more such 

 structures than any other in the country uses solid-plate webs. 

 The writer is accustomed to use a web system for this circular 

 girder composed of angles. Usually this girder is utilized to 



Circular Girder, 

 3'deep 



