How dependent is vortex formation on site specific
geometry and operating conditions?
This is one area where all authors in the literature review agree that the site
specific conditions of the approach flow is one of, if not the key parameter in
determining if a vortex will form. When listing the ideal conditions that pumps
should be designed for, Prosser (1970) recommends a hydraulic model study if
there are any deviations from the stated design guidelines. Anwar (1981) and
Pennino (1980) both advise that it is very difficult to estimate the bulk circulation
without the use of a scale physical model since the circulation is dependent on
many factors in the approach conditions which are entirely site specific. The
approach flow is just as important if not more so than the dimensionless
submergence and Froude number, and “there is virtually no submergence at which
an intake designer can be certain of vortex-free operation” (Pennino, 1980). Bauer
(1997) elaborates on this and states that it is difficult to detect submerged vortices
unless studied in a laboratory that can investigate with dye or air injection.
Figure 3 by Rindels (1983) and the similar work by Gulliver et al. (1986) can give
a rough idea of vortex risk, but they cannot account for approach flow and
geometry that can significantly influence vortex formation. More rigorous
evaluation of vortex risk can only be achieved by construction of a physical
model and visually inspecting for vortices (Suerich-Gulick et al., 2014). It is very
important that the construction of a physical model to investigate vortices match
the geometry and flow approach conditions exactly, as was found with the initial
model of the St. Anthony Falls Lower Lock vortex study. The first model was
constructed at the St. Anthony Falls Hydraulics Laboratory and due to space
constraints the full prototype approach topography was not modeled. Tests on this
model did not observe vortex formation, and lead to the second physical model
study at WES (Ables, 1976).
Most laboratory scale models employ Froude number similitude which match the
Froude number of the model flow to the Froude number of the prototype flow and
allow the dominant forces of gravity and inertia to scale correctly (Jain et al.,
1978). However, with water being the predominate test material, it is impossible
to match the Weber and Reynolds numbers which leads to scale effects of both
surface tension and viscosity (Suerich-Gulick et al., 2014b). Rindels (1983)
recommends that the Reynolds number of scaled model flows be greater than
5x104 to avoid viscous scale effects which is similar to standard guidance for
general Froude scale physical model studies (Hydraulic Laboratory Technique,
1980).
Surface tension scale effects are only applicable when an air core vortex is
observed in a physical model, and is stated that the development from a surface
dimple to an air core would occur more rapidly in the prototype than in a Froude
scale model (Rindels, 1983). Jain et al. (1978) performed research on surface
tension with a water & cepol (carboxyl-methyl cellulose) mix that matched the
12
