open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello there,

We are trying to investigate the problems wave deformation when waves pass over different topologies mainly for intermediate depths. We have tried using Artemis, but some nonlinear effects could not be described properly due to the limitation of Airy wave theory. To better describe the nonlinearity, we have tried the Telemac-2D with the Boussinesq approx. Before carrying out the real task, we first set up a simple straight channel for preliminary testing, and we found this one seems to behave strangely when kh approaches 1.13 and beyond. After some research in the forum, we decided to try Telemac-3D with the non-hydrostatic setting, but the wave height seems to decay rapidly. The questions we would like to ask are:
1)Would it be appropriate to use Telemac-3D with the non-hydrostatic setting for water waves of intermediate depth? Are there any limitations, e.g. limits of kh?
2)If the answer to the above question is yes, how can we avoid such strong decaying in wave heights? Reducing the time step seems to work for the hydrostatic setting, but not the non-hydrostatic setting.
3)Is there any recommendation for, e.g. the number of layers, to improve the stability/accuracy of such case?
The attached files show the comparison of simulated water depths of the hydrostatic and non-hydrostatic settings at x=15 m for the simple straight channel with waves of period = 1 sec, wave height = 0.02 m, and grid size=0.025 m settings (ver. V7p3r1).

Hi
First of all it's better to use the latest version so please upgrade to v8p0
And for the configuration, have a look on the velidation test case "solit" which deals exactly with this kind of problems

Regards

Hello,

Thanks a lot for the suggestion. Finally, we found the test case “NonLinearWave” better suits our needs and we were able to create a decent example for further investigation.

For our future scenario, we need to put an obstacle inside the flume and we would like to reduce possible computational loading and the effects of reflected waves caused by the bilateral boundaries. For this purpose, we tried to change the bilateral boundaries from solid wall to open boundary, given h、U、V (4 4 4) and Thompson boundary condition, and the regular wave boundary condition was given at the beginning of the channel. We noticed the wave crest/trough accumulates towards the middle of the channel, which leads to a higher crest/deeper trough compared with the ordinary observation (shown in the attachments). Does someone have any idea to fix or improve this? Or maybe the reason why such an effect appears?





By the way, we also found something strange, which might be a bug? When running the test case “NonLinearWave”, the exported layer-based velocity seems to be wrong. The exported layer-based velocities of U and W are almost zeros, and V has values, which is not reasonable. While disabling the exported variable of dynamic pressure (DP), U and W have values and V becomes almost zero, which makes more sense.