but not as strongly as exhibited by the higher flow rate (see plots in Appendix C).
In fact, dye injections at the slower flow rate suggested that the jet spreading was
occurring higher in the water column.
As in the previous tests, the velocity component ratios between prototype and
distorted models were determined at each measured point in the flow. Ratios of
the crossflow velocity components for experiments with discharge scale NQ = 1.0
are presented in Figure 44 for elevation 2/3 d above the bottom, and the
corresponding ratios in the principal flow direction are plotted in Figure 45.
Ratio plots for lower in the water column at elevation 1/3 d are shown in
Figures 46 and 47. Complete results are included in Appendix C.
Case 3 discussion and conclusions
Jet flow separation initiated by the sloping edges of gap in the Case 3
experiments was expected to generate turbulence having strong components in
both horizontal and vertical planes. This assured that several of the turbulence-
related terms in the equations of motion would not be in similitude in
geometrically distorted models.
At the higher flow rate the prototype jet exhibited spreading at lower depths
brought about by upward fluid entrainment by the jet. This effect was not as
pronounced nearer the free surface. In the distorted scale models, flow
separation was caused by much steeper gap edges, and this resulted in less
upward fluid entrainment and less spreading of the jet at lower depths. Dye
injection near the jet boundary gave visual confirmation of turbulent spiral-like
flow structures moving in the principal flow direction. The size of the spirals
decreased as model distortion was increased.
The observed scale effect did not have much impact, if any, on the main
nonturbulent region of the flow; and away from the jet boundary, the flow
entrainment seemed to be in reasonable similitude. The main impact of the scale
effect was in the immediate vicinity of the jet boundary, and probably farther
downstream where the jet becomes fully turbulent. The fully-turbulent jet in a
distorted model will probably not be quite as spread out as the prototype at
deeper depths. However, near the free surface, the jet seems to maintain similar
velocities and geometries. The severity of this scale effect in a geometrically
distorted model, and its impact on study results, must be evaluated according to
the specific physical model configuration; but overall, the primary flow
structures do not seem to be affected at any great extent. In other words, some of
the turbulent structure geometry will be incorrect, but the distorted model should
reproduce somewhat similar dominant flow patterns associated with flow
separation with velocity magnitudes nearly correct.
Chapter 5 Turbulence Scale Effects Experiments