Choosing a Computational Increment – Sediment Transport

[JohnLonergan]

Hi,

I am trying to decide how to select the computational increment. Is this equivalent to the time step in an unsteady analysis?

If so is it acceptable to specify the CI for each period of the quasi flow file according the Courant equation in the HEC RAS manual? I appreciate that a different CI is necessary for low and high flows to ensure an efficient model run. Using a formula would be ideal as I could do it with excel and then paste the values into the Quasi flow editor.

If so , to do this would require the value of Vw. Does anyone know how to extract this data from HEC RAS?

Failing this, would it be acceptable to use a small CI of say, 0.1 or 0.2 hours? It may be accurate for high flows but a drawback is that it will ages to run a simulation for 8760 hours.

Any wisdom would be more than welcome.

PS I had other threads running on Sediment Transport problems where it would not work. I have now solved these and it is running but I am just trying to refine the model now.

[Anonymous]

I don’t think that the sediment computation increment is as sensitive to the Courant condition as unsteady flow is. An increment of 0.1 or something sounds to me smaller than you would probably want for most data sets. You could try something like 1 hour and then run it again at 0.5 hours and if the results look the same, the 1 hour was probably ok. Sediment also allows you to vary the time step based on the flows. So for peak flows, the smaller time steps might be reasonable.

[JohnLonergan]

Hi Jarvus,
I will experiment with what you suggested tomorrow.
I will post again with the result and let the community know how this works.

Many thanks for the advice.

[Author]

Posts