Jump to content
GMS, SMS, and WMS User Forum
amy

RMA2 stops: punctual very high velocities

Recommended Posts

Hi,

I am using RMA2 in SMS 11.0.03 for simulating a flood in a river/ estuary. I am facing two problems:

1- In the calibration process, I would like to simulate during 12 days, including 1 spring tide and 1 neap tide. I go from 6m to 2.99m (first value of the tide, starting with neap tide) using the steering and then I use that hotstart for the unsteady simulation (with wet/dry option). The problem is that the simulation allways stop after simulating the first 6 days, i.e., when starting the spring tide. I verifyed that I have abrupt high velocity values in a sand bank (1 or 2 nodes) and because of that Rma2 immediatly stops. I tryed to increase the E viscosity form 10,000 to 30,000 but I stil get problems. Any sugestion? Is it convenient to start by the spring tide cicle?

I am creating hotstarts every 150 iterations. Is it possible to use the iteration, in the hotstart, corresponding to the first e.g. 5 days (the velocities were ok) and then changing the conditions for the remaining days of the simulation e.g. E viscosity, wet/dry versus marsh porosity, time step, etc. I tryed some oh those and the Rma2 stops a few minutes after.

2- For simulating the flood, I have to start form 22m (instead of 6m). The problem is to go from 22m to 2.99m, since the steering allways stops. Any idea?

From your experience, all help and suggestions are wellcome.

Thank you very much.

Amy

Share this post


Link to post
Share on other sites

Amy,

Without seeing the specific geometry of your model it is hard to make to many suggestions, but I had a few thoughts.

1- Spinning down with steering (or manually) to the neap tide is a valid way to start the process. You should make sure that you really are converging when you get this value. RMA2 can continue on when it is really not converged.

2- Do you have any "puddles" forming at low tide? It is a fairly well known stability issue that RMA2 will go unstable when the water level starts to rise and puddles (wet areas that have become disconnected due to drying) start to reconnect to the main wetted area. You can connect theses areas via narrow channels with high mannings n values to try to improve stability.

3- It seems like your Eddy viscosity is really high to begin with. I would be wary of using values that large.

4- RMA2 has never been stable in super critical flow. If you have significant supercritical flow areas I would suggest switching to another model. The army has its new model (AdH) which I have used for tidal circulation for a couple of situations. There is also the ADCIRC and TUFLOW-FV models that handle supercritical flow just fine (both of those require commercial licensing). SMS includes a complete interface for ADCIRC and TUFLOW-FV is available through the generic model interface. I have not tried the SRH-2D model from the BoR for tidal, but it does handle supercritical flow. It is also available through the generic model interface.

5- It is possible to to use any hot start values at any point in the simulation, but it gets tricky and is not a true representation of the conditions. It is sidestepping the instability when it is better to resolve it.

Share this post


Link to post
Share on other sites

Hi Alan! i'm having trouble when using supercritical inflow as boundary condition (SRH-2D). The model won't even run and I receive this error message during the execution of srhpre:

********************************************

* Program stopped due to the following:

* Time_Series data file does not exist!

* Error Code is: 6

********************************************

This error is not in the list of SRH errors so I don't know what to do.

The same mesh works fine with subcritical inflow but when I switch to supercritical it won't run.

Thanks in advance.

 

 

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×