RAS numerical instability issue

[Olkeba]

Hello everyone,

I’m very new with the HEC-RAS model. I successfully ran the model for 1D steady flow of 100 yrs storm, but the model failed when I ran for 1D unsteady flow (see attached error screen shot).

I modified the cross section data, simulation time even to 0.1 and reduced cross section space interval. Nothing resolved my problem and when I modify one cross section, the problem jumps to another cross-section with numerical instability.

I’m wondering if RAS is applicable for steep slope (> 0.04) and short channel length ( <= 200 m). Could anyone in the forum help me to fix my problem? Thank you so much in advance, Olkeba

[Anonymous]

I believe hec ras limitations for slope is 10% or 6% cant remember exactly but its governed by eliminating the pressure component of the unsteady equations (st venant) due to the cos of an angle greater than a certain degree.

Anyways a model at that slope that is short like that is probably just drying out. Read into pilot channels. Add one of those and it probably will solve.

[Olkeba]

Hi Luis,

Thank you for the information. Where should I “read into pilot channels” and “Add one of those”? Sorry, I’m quite new with HEC-RAS.

Best regards,
Olkeba

[Anonymous]

Hit F1 when you have hec ras open. This brings up a manual

[cameron]

You could also set a minimum flow so the cross-sections don’t dry out. What do your htab parameters look like?

[Olkeba]

Cameron and Luis,

Thanks for your suggestion. Will try accordingly.

@cameron, here is the htab parameters. I edited starting EI to invert level, reduced Increment, and increased Points. Any suggestion from there?

Thanks.
Olkeba

[cameron]

Check your ineffective areas. You don’t want the active top width changing between cross-sections to rapidly as that will also cause models to blow up.

[Olkeba]

All the ineffective areas are zero. Are they supposed to different from zeros?

[cameron]

That depends on the situation. You don’t want the active flow area of each cross-section changing significantly from one cross-section to the next as it causes a shock to the system. For unsteady modeling it is a lot more difficult to deal than than steady and can require multiple ineffective areas.

Water cannot expand faster than 2:1 to 3:1 and can’t contract faster than 1:1. This is important around bridges and other areas that expand or contract quickly.

[Olkeba]

Thanks Cameron for the input. I will check for the ineffective areas.

Best,
Olkeba

[Author]

Posts