Instability of model

[Thabi]

Hey!
I have been running a model under the unsteady flow and it consistently gives me the same error, even tho I have already tried all possible solutions given in this forum. The latest error given was:
Maximum iterations of 40 at: RS WSEL ERROR
Solution solver went unstable, iteration 40 at 20JAN2014 12:59:00
Gave_Oloron Gave_Oloron 57846 145.74 16.950

**** Solution Solver Failed ****

Minimum error exceeds allowable tolerance at 20JAN2014 1259

Gave_Oloron Gave_Oloron 62222
WARNING!

***** Extrapolated above Cross Section Table at: *****

Gave_Oloron Gave_Oloron R.S. 68562
Gave_Oloron Gave_Oloron from R.S. 68033.* to 67764
Gave_Oloron Gave_Oloron R.S. 59529
Gave_Oloron Gave_Oloron R.S. 59219.*
Gave_Oloron Gave_Oloron from R.S. 57821 to 54909

After this, the red bar starts running so I am kinda blocked on how to proceed to correct these errors. I have already increased the number of stations and limited the spacing to 300 m, I have changed the increment to 0.25 and the number of points to 100 – even tho it still extrapolates. Any help or suggestions on how I can fix it? Thanks

[cameron]

Have you tried adjusting the time step? Basically the model crashed for some reason causing it to calculate a very high water surface elevation that is not real which then produces the extrapolation error.

[Thabi]

I have tried (almost) everything! I have taken different time steps and it still crashes. I have readjusted the cross sections also (based on the warnings given in the outputs), but it doesnt really solve anything. I used the initial conditions as the first point of the hydrograph (even tried changing the ranges) and it still gave me the errors. No mater what I do, the simulation crashed!

This is the result of the last simulation I did:

[Thabi]

I have gone through nearly all the possibles reasons why a simulation can crash, I have changed as much as I could and still NADA!

[Anonymous]

Perhaps this is a critical flow issue. Do you expect the flow to be supercritical anywhere? Maybe using the mixed flow option will get you started.

[Thabi]

I have also tried using the mixed flow condition, and still, unstable :/. I ran out of options and I dont know what else to do! If anyone can please give a second look at the model https://drive.google.com/open?id=0B8YfqX_u7Oqnc0tXUGp0dkdLdXc
Thanks

[cameron]

I took a quick look and here are my initial thoughts. Your cross-sections are too far apart at a lot of locations especially near the many drop structures you have. The HEC-RAS manual has guidance on how far about cross-sections are supposed to be. I interpolated cross-sections every couple hundred meters and tested it the model and it appears the drop structures are giving the model a hard time. I would also think about changing the HTAB parameters so that it starts at your invert and have a decent elevation change of .1 meters or someting better than 0.3 or higher.

I would say a 2D model would be more stable, but the model is long and your hydrograph is for a very long time so run times might not be worth it.

[Thabi]

Hey Cameron. Thank you so much for taking the time to see my model. About the cross sections, that was also my initial thought, therefore I did add more spaced at about 200 m from each other (I zill try reducing this spacing), and I have also considered that the amount of points in the hydrograph is very high – I did reduce to 100 points and it still became unstable! I did put all the starting elevations at the invert and changed the intervals. I checked the output files and I have negative discharges at some random points and also near the inline structures. What can I do about the drop structures? Running the model on mixed flow – which is what the results show that happens – does not really yield any proper results. I think I am super stuck!

[Thabi]

I have tried the simulation without the inline structures and now the model does not become unstable – there are errors of course that I am trying to work on regarding the iterations and stuff. Any suggestions on how to work on them?

[Author]

Posts