f.
Flow separation at a vertical step where the turbulence is manifested
primarily in the vertical plane will not have any significant scale effect
in geometrically distorted models.
g. Steeper slopes in the distorted model will tend to generate less vertical
fluid motion. Consequently, a scale effect will occur between a
prototype where flow turbulence is generated at a sloping boundary.
However, the scale effect is strongest near the bottom and appears to
lessen closer to the free surface. Also, the scale effect seemed to be
restricted to the immediate vicinity of the jet boundary; however,
potential impacts farther downstream were not evaluated due to the
limited measurement region of the experiments.
Importance of turbulent scale effects in a geometrically distorted physical
model relates directly to the processes being studied, and whether or not flow
turbulence is a dominant forcing of that problem. Specifically, the Alaska
District is considering construction of a large physical model of Cook Inlet with
a geometric distortion of 4 (horizontal/vertical length scale), and the Alaska
District engineers are concerned about potential turbulence scale effects because
flow separation and gyre formation are known to be important influences at Cook
Inlet.
Based on (a) performance of the 3-D flow table model with horizontal-to-
vertical distortion of 15, (b) theoretical analyses of potential scale effects, and
(c) flow table experiments, it is the opinion of the report authors that turbulent
scale effects in the proposed distorted physical model of Cook Inlet would not
significantly influence model results. In other words, hydrodynamic flow
patterns, regions of flow separation, generation of large-scale gyres, and results
of sediment tracer and dye injection experiments from the distorted model would
closely resemble those of an undistorted model. Differences will occur in the
differences are expected to be localized and should not influence overall flow
patterns. Three-dimensional flow structures will also have a scale effect with
vertical velocity components being stronger than they should be. However, this
is partially offset by steepening of slopes in the distorted model that will decrease
the vertical turbulent fluctuations. An important consideration in large distorted
physical models is providing sufficient bottom and boundary roughness to assure
a fully turbulent boundary layer.
Geometrically distorted physical models retain all of the nonlinearities
inherent in complex flow situations, including turbulence generation. Even
though known scale effects will alter the value of some turbulence terms, the
overall turbulence flow field will still exhibit most of the behavior expected from
an undistorted physical model. For cases where turbulence and 3-D flow patterns
dominate, the distorted physical model provides more reality than depth-averaged
numerical models or even 3-D numerical flow models which discard or linearize
the convective accelerations and/or simplify terms related to turbulence.
Consequently, where space limitations prohibit construction of an undistorted
model, geometrically distorted physical model may be the best tool for
examining engineering problems related to flow turbulence and 3-D flow fields.
92
Chapter 8 Summary and Conclusions
Previous Page
