open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I've made a thin canal with a length of 100m, a width of 1m and a depth of 1m. I have a prescribed a flow rate of 10 on one side and a prescribed elevation of -0.5m on the other side in order to allow water to flow out.

Upon running a simulation, the water depth increases to unrealistic values of 20m and I cannot seem to identify the problem. I have attached the relevant files.

Hello,

In my view, this is only a matter of setting of your problem.

Ponder about it: You have a input flowrate of 10 m³/s in a section area of 0.9 m² (assuming the free surface is at z=0). which yields an input velocity of ~11 m/s. The at the outlet, assuming ideal flow, if you have a water height of 0.5 m, then the section area would be 0.425 m² and your velocities would have to be twice as much as in the inlet. This is the idealized case.

In the simulation, you also start with an initially dry domain, that takes a little while to become wet and give a stable flowrate for your input. This fact only makes your free surface at the inlet go up more than you would expect to. For this I would suggest you impose an initial elevation of 1 m along your domain, so the water would flow faster.

But what I see as the main cause of the high elevation is your bottom friction setting.
You are using Manning's law with a coefficient of n = 4. If I made the calculations correctly, with a height of 0.5 m at the outlet, and 21 m at the inlet, your channel would have a discharge velocity less than 1 m/s, and remember that we calculated an outlet velocity of ~22 m/s. In this case your channel is not letting the water out and your surface at the inlet is rising until it reaches an steady state (that will be far more than 21 m height).

I would suggest checking your friction law and coefficient. Maybe you used one law with the coefficient from another... (Manning's n is 0.014 for concrete, for example. Yours is far too high).

Another thing that I suggest is increasing the number of points in the along-channel direction. this case with inlet height of 21 m and outlet height of 0.5 m is causing instabilities due to the distance between the points.

Regards,

Phelype

Dear Phelype,

Thank you for your reply.

the suggestions you have made have greatly improved my simulation.

I have set the mannings co-efficient to 0.03. The initial elevation on the steering file is set to 1. The distance between each node is 10 times less.

The volume at end is 312 metres and the original volume is 100 metres and it takes 20 seconds for the flowrate to reach the end and give a flowrate of 10ms. I don't know whether I should add an incline to the river bed in my geometry file to improve accuracy.

Hello,

It's not a matter of accuracy. You can simulate a horizontal river bed if what your are trying to simulate is horizontal. It's up to you to decide if it should be done or not.

What can (and probably will) improve your results is solving that message telemac is giving you:This means that the conditions you impose on your system lead to an infinite number of solutions, which can lead to impossible scenarios. For example, I ran your simulation here (old mesh, new configuration) and at the inlet, there is a node with free surface of -1.02 m (below the bottom!) right beside a node with 2.45 m. They are 10 cm apart.

This happens because there is not enough boundary conditions to make the solutions of your model unique. There is a lot of discussion on this forum about this topic. Just google "ILL-POSED PROBLEM site:opentelemac.org" that you will find a lot of better explanations about the problem and possible solutions.

Regards,

Phelype