restricted to studies o£ tides, circulation patterns, salinity intrusion, 

 and hurricane surges in broad, unrestricted waterways, a distortion ratio 

 as high as 20:1 can be used. However, in past models with such a distor- 

 tion, it was found extremely difficult to achieve a satisfactory repro- 

 duction of vertical velocity and salinity distributions in well-defined 

 channels. 



After the vertical and horizontal scales have been selected, it is 

 necessary to ensure that the flow in the model will be turbulent through- 

 out most of the tidal cycle. Turbulent flows will exist when the Reynolds 

 nuniber, dV/v, for the model is on the order of 1400. In most instances 

 the vertical scale should be greater than 1:150 to satisfy this criterion. 

 This requires that conqjrehensive estuary models be built to distorted 

 scales, since it is normally economically infeasible to use a horizontal 

 scale greater than about 1:300 (the largest feasible scale is usually on 

 the order of 1:500 to 1:1,000). 



c. Scale Effects . There are several significant scale effects common 

 to estuary models. The distortion of linear scales directly influences 

 the required model roughness (eqs. 3-8 and 3-10). Since the roughness 

 ratio increases with the distortion ratio, models with high distortions 

 may require so much artificial roughness that the flow regimen is severely 

 disrupted throughout the vertical to achieve the proper degree of mixing. 

 However, it is not as significant a factor in broad, shallow bays with a 

 small tidal range where flows are naturally low and mixing is primarily 

 generated by wind. 



The flow through an inlet, canal, or structure is dependent on its 

 resistance (roughness) characteristics. The reliability of results of 

 tests on proposed openings of this nature will depend on ensuring that 

 their flow characteristics are properly modeled. For this reason it is 

 usually necessary to conduct flow calibration tests in an undistorted- 

 scale model (usually a flume-type facility). After the flow character- 

 istics have been determined in the iindistorted-scale model, the opening 

 is subjected to calibration tests in a distorted-scale model (still in 

 the flume) during which its width or shape is altered to produce the 

 required flow characteristics. The final configuration is then con- 

 structed in the distorted-scale estuary model. (A detailed discussion 

 of undistorted-scale modeling is presented in Section VII.) 



A similar scale-effects problem is encountered for flow through pile 

 structures (e.g., pile dikes). If the structure is modeled according to 

 the horizontal scale, the openings between piles are likely to be so 

 small that surface tension will adversely affect flow through the struc- 

 ture. Previous flume tests have determined that the vertical scale should 

 be used in modeling the horizontal dimensions of such structures. 



Short -period wave action cannot be accurately reproduced with the 

 distortion ratios commonly used in estuary models. Short -period waves 

 are generated in the models only to simulate the effects of wave 

 energy on the resuspension and movement of sediments in the model. 



59 



