Jacob H. Douma
This would add a lot to the cost because a higher wall costs more, and the variable wall
height makes construction much more difficult. So we got the idea of designing the
channel like a high-speed highway with spiral transitions between straight sections and
curves. The spiral transitions gradually change the direction from a straight line to a
curve. After passing through the curve, a reverse spiral is used to gradually change back
to a straight line. In addition, highway pavements are super-elevated around sharp curves
to provide a gravity force component which improves the passage of high-speed traffic
around a curve.
We reasoned that flow in a high-velocity channel would be similar to traffic on a
speed highway. The
scale model was reconstructed with super-elevated curves and
spiral transitions. It was remarkable how much better the water went around the curves.
Flow depths were essentially equal in the curves along the inside and outside walls, which
made it possible for those walls to be equal in height. Also, the super-elevated transitions
reduced wave disturbances which otherwise occurred because of sudden changes in flow
direction of the water.
We measured five feet high waves on the outer side of the curve. The waves progressed
quite a distance downstream from the curve. Higher walls would be required to take care
of those waves. By super-elevating and using spiral transitions in the curves, these waves
were greatly reduced in the
scale model. We decided that this was a very significant
improvement that should be tested in a large-scale model.
Fortunately, the outlet works and spillway were already constructed at Hansen Dam. The
outlet works discharged on the spillway chute, which was quite flat. A flume was
constructed at the end of one of those outlets, in which a 1: 10 scale model of the
rectangular Tujunga Wash channel was tested. This model was tested with and without
spiral transition curves and super-elevated inverts to compare it with the
scale model.
The 1: 10 scale model generally confirmed the previous
scale model test results. Only
small height waves formed in the spiral transitions and super-elevated curves. Also, the
depths of flow were essentially the same through curves as in adjacent straight channel
sections. That reduced wall heights and construction costs appreciably. The channel was
designed and constructed that way.
We made a great point of spreading this information throughout the Corps. We said,
"This is the way
velocity channels should be designed." That's the way they've
been designed ever since by the Corps and others who have designed high-velocity
channels. I don't know how much it has saved over the years, but certainly it saved a
great deal of money.
Previous Page
