teristics of ice movement through confluences. The significance of selected key
parameters was investigated using the diagnostic model.
Confluence bathymetry
The influence on ice movement of confluence bathymetry can be inferred to a
certain extent from the parameters in eq 15 and 20. As mentioned in the preceding
section, the size of the bar approximates the size of the flow separation zone, though
actually it may be slightly larger. Therefore, as an approximation, bar presence
and its influence on ice movement are reflected partially in the magnitude of
bc/b3. Bar presence, however, also is associated with substantial variations in flow
depth through the confluence.
The shallow depths of flow around the bar have several adverse effects on ice
movement through the confluence. Ice may ground in shallow water. The speed
of ice movement in shallow water is commensurately reduced with shallower flow.
Also, during periods of frigid weather, border ice may form rapidly from the edge
of the bar, thereby even further reducing the surface area of flow available for ice
movement.
JAM PROCESS MODELING
A small-scale model was used to confirm the importance of the key parameters
identified in eq 15 and 20, and to establish the principal processes limiting ice
conveyance and leading to ice jam formation in confluences. Particular attention
was given to determining influences on ice conveyance of flow and bathymetry
features at confluences of concordant channels. Since the modeling required numer-
ous alterations in channel orientation, size, and bathymetry, as well as variations
in water and model ice flow rates, the model was kept small in overall size, flexi-
The findings produced by the model are largely qualitative. They essentially
comprise the description of two principal processes that limit ice movement through
concordant channel confluences and lead to ice jam formation. The two processes
are formulated in a subsequent section of this report. Both formulations can be
used to quantify the limiting rates of ice movement through confluences.
Model setup
The setup for the process model comprised the basic components of an elemen-
tal confluence of two concordant bed channels. The channels were rectangular
in cross section and were built so that the angle between the two inflow channels
could be adjusted. Channel widths were adjustable, but bed elevation was held
constant for all the channels. Figure 10 shows the model setup, which also is
depicted by the photograph in Figure 11.
Flow to the inflow channels discharged from a head box at the upstream end of
each channel. Flow from the downstream confluent channel passed into a tail tank,
from where it was recirculated by means of a pump to each head box. The flow
rate to each inflow channel was controlled using a valve in each supply line and
measured using orifice plate meters. Each confluent channel was sufficiently long
that more or less uniform flow developed in the channel. Observation of water
surface elevations confirmed that the flows were acceptably uniform. Flow
visualization with dye confirmed that velocity profiles throughout the greater
length of each channel conformed with steady, uniform flow in each channel. The
22
Go to contents page