open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi there,

I would like to know if there is a way to prevent flow from entering a boundary with prescribed H value. I have a very large boundary located on a very shallow weir. I impose a stage-discharge relationship, but flow is entering the model from approximately half of the boundary, creating a large recirculating flow in front of the weir, which has obviously no physical meaning... (there is no weir in my model: it is replaced by a boundary)

Thanks!

Hi
Why you didn't impose directly the discharge?
Imposing only H on an entry which is subcritical gives an ill-posed problem.

Hope this helps

My boundary is not a upstream boundary, its a downstream boundary. But Telemac compute entering flow on half of its length.

Okay... I mean "prevent" not "prescribe". Sorry for the confusion...

Since an image worth 1000 words...

Hello,

I am not sure that what you do is correct. A very shallow weir in the exit could give a transcritical flow, then the condition on depth should be relaxed otherwise you could see strange effects, like velocities entering because the prescribed elevation at the exit creates a step... Anyway I would start first to establish a steady state with a constant free surface elevation at the exit. For example if you start with constant elevation in your domain, and a slightly lower elevation at the exit, the flow will clearly go out. Then if it works you can start applying the stage discharge curve, with relaxation so that it evolves smoothly.

Regards,

Jean-Michel Hervouet

Hello Jean-Michel,

thanks a lot for this explanation. I've already did what you proposed (initialisation with constant depth at the boundary and initial water level slightly higher). I thought I did get the steady state. The image I'm showing is a continuation of the simulation with a rating curve trying to find the steady state with that boundary type (I need the rating curve because the next step is a dam break upstream). If I let the model run for a long period of time, than it gets back to what it seems correct answers from the model.

By the way, the downstream weir itself (high bottom elevation) is NOT included in the model (only its rating curve). But the bathymetry in front of the weir is high, so that could explain it.

I would like to ask you to define a little bit more what you refer to when you say "applying the stage discharge curve ... with relaxation". What would be "relaxation"? Is it an option I can turn on? How? In what circumstances it is appropriate to use it?

Many thanks!

Hello,

Relaxation : look at the last line of function sta_dis_cur (library telemac2d). Coefficient 0.02 is the relaxation, it should be 1. to get a level instantaneously corresponding to the stage-discharge curve, but if you do so it can trigger resonances or uncontrolable oscillations (because of the feedback of the level on the local discharge). So we delay the process.

Regards,

JMH