open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I was trying to model a river using "Tubes" and I encountered the following problem : the tube seems to always be in pressurized flow.

* lets consider a river stream which can be split in two parts : one part upstream a bridge, the other part downstream this bridge, the bridge section is modelled using a tube (as is the T2D bridge example)

* the channel width is 5m, its banks are at 3H/2V, and the river stream slope is 1%.

Here is my problem :
* Whatever are the discharge at the upstream boundary and the tube dimensions (width and height) the channel upstream the tube is progressively filled by water so that the tube is "pressurized" in order to transfer the incoming upstream discharge in the channel downstream the bridge.

As attached files I put the t2d model i set up.

According to me, the tube geometry is largely sufficient to transfers 8m3/s (L=6m, H=4m is quite a big tube!), without any "pressurized" flow.

I also noted that when the mesh size is decreasing the water level at the tube downstream node is much higher than water level of the neighboring nodes ... i tried to decrease the time step but I did not see any improvement, (i also tried to play with the FREE SURFACE GRADIENT COMPATIBILITY)... any clue about this?


Thank's in advance for your help.

Best regards,

Thibault