Depths in the model were idealized by three horizontal surfaces located
at elevations corresponding to depths of el 0, el -30, and el -60 with vertical
transitions between the three depths. The glass bottom of the flow table served
as the el -60 contour. Figure 14 shows the small-area model being cut out of
3.81-cm (1.5-in.-) thick Plexiglas, and Figure 15 illustrates the idealized
bathymetry in the completed small-area model as viewed from the seaward end.
Contours are shown on Figure 16 where the red lines represent the shoreline,
the blue lines are the el 0, and the green lines are the el -30 depths. As
mentioned, the bottom of the flow table represents the el -60 depth. The nearly
right-angle bend in the modeled region resulted in the model being built at a size
smaller than originally intended to allow a more direct inflow path while still
accommodating outflow. The model was designed to be reversed on the table for
ebb and flood conditions as illustrated in Figure 17. Operating procedures for the
small-area model are the same as described for the large-area model.
Small-Area Model Observations
The small-area model was also tested extensively during the May 2002 visit
by Alaska District engineers. Flow patterns recognized in the large-area model
were replicated in the small-area model to see if the more detailed (but still
idealized) bathymetry altered any of the flow patterns substantially.
The small area model produced similar flow patterns as the large-area model,
particularly the three large flow separation and eddy formation features at Point
Woronzof, Point MacKenzie, and Cairn Point. The bottom crosscurrent flowing
diagonally between Point Woronzof and Port MacKenzie was also evident.
The small-area model did not replicate the boundaries as far upstream as the
large-scale model, and tests showed that changes to these boundaries could have
a downstream impact. It was concluded that care must be taken to reproduce
accurate upstream boundaries in flow table models so that flow patterns are
reasonable facsimiles of the actual region being modeled. This was an obvious
constraint in using the small-area model.
The effect of vertical transitions between horizontal layers was investigated
by using modeling putty to place a fairing along the vertical transition between
the layers. This gave a sloping transition, which allowed vertical mixing to
occur between layers. Thus, the main drawback to the idealized bathymetry is
prevention of vertical exchange between depths. This becomes a problem only
for cases that where 3-D flow structures with non-negligible vertical flow
components are known to exist.
Conclusions From Idealized Models
The two idealized physical models of Cook Inlet proved to be valuable in
understanding the complex flow patterns in the vicinity of Anchorage, AK. Even
though the models were very small, constructed at a highly distorted geometric
scale, and represented fairly complex bathymetry by simple horizontal terraces,
Chapter 3 Idealized Cook Inlet Models