open TELEMAC-MASCARET The mathematically superior suite of solvers

Try with a higher initial water height (with higher downstream boundary condition) and reduce it smoothly.

regards

This is probably due to the existance of supercritical area at the entry. The problem will be ill-posed if this is the case. You shoud
- change the bathymetry at the entry area in ordre to decrease velocities
- or/and prescribe an elevation in addition to the flowrate and work with 5 5 5 condition at the inlet boundary.


regards
Riadh

Hi,

with Method of Characteristics for velocities try to use SUPG = 0;1. It gives you fully upwinding of the depth which is beneficial in the case of higher Froude numbers. Otherwise sometimes I encountered the same issues even if the Froude number is well below 1.0.


Best regards,
Clemens

I have tried with the higher outlet elevation and low it. it is doesn|t work.


|supercritical area|

In the experiment, the velocity is quit smooth in inlet.

For simulation, I maybe of cource.

I think it is to change these areas near inlet bottom elevation, and set the

the boundary condition as follow figure. Am I right ?, I will try it now.

It is still the same.

and when I set the velocity and elevation at inlet. the velocity in the whole canal become high unaccuracy in this time.

SUPG = 0;1

it is give the result like following.





I will also try to do simulate a long time, maybe the results will be better.

But it is inspired me, that this is maybe due to the problem of the dicreted scheme

Ok, now I tried it (version v7p0r0). Taking your cas file and setting SUPG to 0;1 gives a steady state flow field.

N.B. SUPG 1;1 or SUPG 2;1 gives the same result since the Method of Characteristics doesn't use SUPG upwinding.

A question to the developers:
sometimes with quite simple geometries and hydraulic boundary conditions I run also in this issues like Harryyao and the only countermeasure seems setting SUPG upwinding for depth like 1.
However other advection schemes for velocities like 14 will fail since they don't allow depth upwinding with SUPG. This happened in some projects when the Froude number was high like in dam-break cases and so on. Is there a better way to stabilize the simulation?

Best regards,
Clemens

I try SUPG=1;1
&
2:2 the unenven velocity distribution in inlet is disapper but it is doesn't give reasonable velocity distribution at all.

I think the question of discrete scheme and difination of the FRICTION COEFFICIENT.

Dear Clements,

I also run it but in V6P3. Set the SUPG 0;1

When you look inside the velocity. for example, set the velocity show range in 0.45 - 0.65 and you will found the velocity is unreasonable.

I didn't try it in V7P0, So I don't know it is difference or not.

Best Regards

For highly intransient flows including rapid wetting and drying finite volumes performs very well. with Finite elements we need to be careful with the option related to numercial schemes.

with my best regards
Riadh