Negative flow at the sluice gate/unknown water source at the gate

[Anonymous]

Hi fellow modellers,

have been puzzled with this problem for a couple of days. I have a 2D model of an area with a dam (with 2 gates) in the middle of this area. I made a time series of opening for these 2 gates. After the model run (with no instabilities and converge problems). I found that there is a negative flow at one of the gates. Due to these negative flow values, the problem gate was sending water from downstream to upstream, even though the water from the model Upstream boundary condition did not reach the dam yet.
Interestingly raising the Gate invert (from 368 meters to 370 meters) did not result in negative flow. I chose 370 because the second gate had an invert of 370 meters. I checked that the elevation of HW and TW Cells, they are not higher than the invert elevation, I’ve given. I don’t wanna leave the invert at 370 meters, because the real elevation is 368.
Does anyone have an idea how this could be solved?
I can not understand how the invert is resulting in negative flow. Or to put it this way, how can invert create water in HEC-RAS?
Hope someone has an answer for this

[cameron]

Did you try turning on flap gates so there can be no negative flow?

[Anonymous]

Hi Cameron,
thanks for your reply. Yes, I did try with flap gates (with no negative flow). But no influence.
Its weird to see water at the gates, considering that I have no water in my model in the beginning and that the inflow water hasn’t reached the gates yet. How can HEC-RAS create water at such hydraulic structure out of nowhere?
I had a look on the HEC-RAS Hydraulic reference manual and the equations are quite normal (for free flow Q=CWB*SQRT(2gH)) and for not free flow (Q=CWB*SQRT(2g3H)). The only difference being the calculation of head difference in these two cases and multiplication by 3. I can’t seem to understand how HEC-RAS calculates negative number, if I know that TW side is lower than the HW side and there should be no water. MY mesh size is 50 meters (I know it might be big, but my overall territory is huge). I tried to play with computation interval from 1 minute down to 10 seconds, but no influence again.
Any other suggestions would be highly appreciated

[cameron]

for your gate invert values, are they below the lowest elevation of the grid cells they are connected to?

[Anonymous]

they are not lower than the HW and TW side cell elevations.

But luckily I found the reason, at least for now there is no negative flow. My structure was between two 2d flow areas. So when I played around with the perimeters of my 2d flow areas, making them smoother and created the mesh again, the negative flow disappeared.
I’ll post if the problem will reappear.

[Author]

Posts