RE: Boundary conditions for a man-made pond - Home - Delft3D
RE: Boundary conditions for a man-made pond
I am relatively inexperienced in FLOW and as a result have been unable to produce what I believe to be realistic flows fields for my model of a man-made waste water pond.
I have attached a schematic describing the scenario. Flow enters through a pipe at an average rate of 0.023 m3/s (on left). It then exits the pond through a weir (on right) where the Q-H relationship is:
flow = 0, H = 0
flow = 0.023 m3/s H = 0.065 m
where H is height of water above the weir. I have tried the following configurations for my flow boundaries/discharges:
Total discharge time series boundary (0.023 m3/s) in and Q-H water level boundary out (flow = 0, H = 0; flow = -0.023 m3/s H = 0.065 m)
Discharge operation at inlet (0.023 m3/s) and discharge operation at outlet (-0.023 m3/s)
Discharge operation at inlet (0.023 m3/s) and Q-H water level boundary at outlet (flow = 0, H = 0; flow = -0.023 m3/s H = 0.065 m)
None of these seem to produce a flow field where water enters the pond and is channeled out at the outlet. Can anyone advise me as to what the boundaries should look like? Perhaps am I making a mistake in terms of defining my flows as positive/negative when they shouldn't be?
I was also wondering if I should be setting my initial water level to 0.065m or 0. In real life water is continuously flowing over the weir. Since I want to start my model with flow over the weir, my thinking is that I want my I.C. for water level to be greater than zero rather than waiting for the water level to rise before flow over the weir and out of the pond commences. Does this make sense?
For open boundaries (e.g. total discharge and Q-H) the sign convention for flow depends on the grid indexing. i.e. if your grid indexing increases from left to right (west to east) and you wanted flow from left to right, then boundaries would have a positive flow. (i.e. flow in the direction of the grid orientation is positive, flow opposite to the grid orientation is negative). Generally (but not always) the indexing increases from left to right and bottom to top so I would expect your Q-H boundary should have:
Q = 0, H = 0
Q = 0.023, H = 0.065
it may also be good to have a higher flow point on your Q-H curve for stability e.g.
Q = 0.046, H = 0.13
Alternatively you could use a Time-series vs head downstream boundary and set a constant head of 0.065.
There are advantages and disadvantages of a total discharge vs a discharge operation but a discharge operation may be more stable (the advantage of a total discharge is you can get flow distributed across several cells.
Hope that helps,
I usually monitor flow at the downstream boundary by inserting a monitoring cross-section into the model at the same location as the boundary - that way flow will be recorded in the tri-hist file. You can insert a cross-section using the GUI under the "monitoring" tab.