306 patterson-simmons. MODEL STUDIES [Ch. 17 



ware River between Artificial Island on the downstream end and 

 Bellevue on the upstream end, plus the tidal reaches of Christina 

 River and Brandywine Creek and about 3 miles of the Chesapeake 

 and Delaware Canal. The model was of the fixed-bed type, molded 

 throughout in concrete to scale ratios of 1:800 horizontally and 1:80 

 vertically. Tides were reproduced by the movement at either end of 

 the model of gates automatically controlled by an electromechanical 

 system designed for that purpose; the roughness of the model channel 

 was so adjusted that the tides and current velocities and directions 

 were reproduced accurately throughout. Shoaling was reproduced by 

 injecting into the model measured volumes of a mixture of water and 

 finely ground gilsonite, a lightweight bituminous material (specific 

 gravity 1.035) which was found to have characteristics most suitable 

 for purposes of the model. The shoaling material was injected into 

 the model upstream from the problem area during the ebb-tide phase 

 and downstream from the problem area during flood tide for several 

 successive tidal cycles. The gilsonite moved from the points of in- 

 jection to and throughout the problem area in suspension and in some 

 degree along the bed; that is, most of the silt was deposited directly 

 in the problem area, while a smaller amount was deposited elsewhere 

 and moved into the problem area by transport along the model bed. 

 After completion of a shoaling test on the model, the model was 

 pooled, and the gilsonite that had been deposited in the channels was 

 picked up and measured. No attempt was made to reproduce the 

 actual volumes of prototype shoaling, inasmuch as the formation of 

 deep deposits of gilsonite on the concrete bed of the model in areas 

 exposed to the tidal currents would have been impossible. Instead the 

 characteristics of the mixture and the method of introduction were 

 varied through a cut-and-try process until the distribution of shoaling 

 material throughout the problem area conformed to the measured 

 distribution of shoaling in the prototype. Taken together, the repro- 

 duction of hydraulic elements, followed by reproduction of the distri- 

 bution of shoaling, comprised the complete adjustment of the model. 

 The operating procedure developed during adjustment of the model 

 — that is, the gilsonite-water mixture used, the volume introduced, the 

 method of introduction, and the exact number of cycles elapsing after 

 introduction of the material — was followed strictly during all tests of 

 improvement plans. In order to assess correctly the effects of each 

 proposed improvement plan on shoaling of the channels, a base test 

 or test of existing prototype conditions was first recorded, and the 

 measured amount of shoaling obtained during each test of an improve- 

 ment plan was compared with that obtained during the base test. 



