HEC-RAS unsteady sim errors

[firenze13574]

Hi all - hope the guys doing the Ohio river sim are reading this

I'm trying to get working an unsteady simulation model of a 130 km river reach that drops about 900 m. The flow data is given in daily intervals for 100 days and intially set the computation interval as 1 day.

At the moment it goes unstable within a few days of simulation-time. I have tried reducing the computation interval to no avail. The error log has lots of warnings such as:

"Warning: The velocity head has changed by more than 0.5 ft (0.15 m). This may indicate the need for additional cross sections.
Warning: The energy loss was greater than 1.0 ft (0.3 m). between the current and previous cross section. This may indicate the need for additional cross sections."

How seriously should take these error messages? I've got XS at 2 km intervals, meaning that the overall gradient between XS is 0.007. If I were to put in enough cross sections so that the energy loss between them is less than 0.3 m I would need on average one every 42 m! Is this really necessary? Or is there something else I should be checking?

Thanks,

Matty

[jarvus lunt]

Are you performing unsteady computations?

"Warning: The velocity head has changed by more than 0.5 ft (0.15 m). This may indicate the need for additional cross sections.
Warning: The energy loss was greater than 1.0 ft (0.3 m). between the current and previous cross section. This may indicate the need for additional cross sections."

This is a steady flow computation warning. After performing the unsteady computations, RAS will perform a post process [steady flow] analysis on select profiles for more detailed output. So if this is what is going on: if the unsteady flow run goes unstable, the post process [steady] results are meaningless and the steady flow errors should be ignored.

If the unsteady flow model is unstable, you might try a shorter time step and examine the hydrograph output to try and determine why/where the unsteady is going unstable.

2km intervals may or may not be sufficient. If the cross sections are large and/or the cross sections do not change radically from one to the next, 2km may be more than sufficient (sounds like it probably is). If there are large changes in area, than 2km may be too far apart regardless of the ground slope.

[firenze13574]

Hi Jarvus

Those error messages are what I see when running unsteady calc (icon of guy running over hill), it fails, and I click on View > Summary Err, Warn, Notes.

Looking at the finished computations window it says:


" **** Matrix Solution Failed ****

Minimum error exceeds allowable tolerance at 01JAN2008 1200

River upper 154.*


Finished Unsteady Flow Simulation"

Is the hydrograph output the same as the view profiles button? That shows the EG line reaching 3,000,000 m over one XS (terrain El. is 2500 m and under). That suggests a problem (!)

I'll try putting in more XS. The terrain is mountainous so relief does change between them.


Thanks for the reply!

Neil

[jarvus lunt]

The summary errors is from the steady post processor run. I can see that that is a bit confusing. Once the unsteady flow has gone unstable, those errors are not going to tell you anything useful.

I sometimes look at the profiles, but more often I look at the stage and flow hydrographs. Which has its own button or you can get to from the view menu.

The key question is to find where its first starts going unstable. Either the wsel is bouncing around or the flows start to go crazy. And then try and figure out what is happening at that location (water surface dropping below critical depth, too much lateral outflow, etc)

[ahsansattar]

Hi firenze!

I am encountering a similar problem as you were having for my 23km river in hilly area.


" **** Matrix Solution Failed ****

Minimum error exceeds allowable tolerance at 01JAN2008 1200

River upper 154.*


Finished Unsteady Flow Simulation"

I want to ask you that how you managed to remove those errors. I also think that I should add more sections. Can you tell me what should be the criterion for distance between the sections.

Thanks in advance.

[ahsansattar]

Hi! Mr. firenze

I am also encountering the same massages for my 23km river in hilly area.

" **** Matrix Solution Failed ****

Minimum error exceeds allowable tolerance at 01JAN2008 1200

River upper 154.*


Finished Unsteady Flow Simulation"

Can you please tell me how you overcome this error?
Can you guide me about the distance between two sections.

Thanks in advance