open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi everyone.
I set the initial state in BlueKenue : a selafin file for the free surface position at t=0s.
I know that the 'CONDITIONS INITIALES' must be set as 'PARTICULIERES' and the CONDIN.f should be modified and introduced in the steering file. Please could you explain to me how this fortran file must be changed according to the initial flow state I've introduced.
Any help is welcome
Thank you in advance.

Finally it work once I use my selfin file as precedent computation (FICHIER DE CALCUL PRECEDENT) and turn on (SUITE DE CALCUL). However TELEMAC2D set (U,V)=(0,0) everywhere which is incorrect cause at t=0s free surface drops upstream to downstream. I even turned on the key word VELOCITY PROFIL (PROFIL DES VITESSES) but stil telemac dose not compute velocity at t=0.
Anyway, even the initial state of the flow is incorrect, I think a long run should naturally serve to get that right initial condition. However still I do not understand why telemac cannot do it.
Please look to the attached, you can find the GEO, CLI, LIQUID, free surface (t=0s) and CAS files and also the steering file.

This is a sample case : rectangular channel (long = 100m, width = 20m, Bottom slope sb=0.001 (bottom drops from 0.1m to 0m). in the selafin file of free surface (t=0s) I set up the water depth at 2m everywhere. I know I can use constant water depth as key word but this is just a sample case before to set up a complex river DEM and boundary conditions.

I ran a steady-flow simulation (Q=130 m3/s). With the indicated Strickler coefficient, rectangular coefficient and the bottom slpe, Strickler-Manning formula profides a water depth h=2m. At the end of the simulation, telemac2d find 2.28m depth rather than 2m.
Thank you in advance.

The correct attached by there
Q=constant=130 m3/s upstream and water level =constant = 2m downstream
initial condition : water depth = 2m everywhere (giving by a selafin file : NEW_TEST_Z_WL_0.slf).
Sb=0.001 → Upstream : DEM=0.1m and Downstream DEM=0m
Channel length = 100m and Channel width = 40m
Circle island → center (50,20) and Radius = 7.5m

Results file → R2D_NEW_TEST.slf

Hi
If you want to initialize the velocity field, you should put the values in the previous computation file like the free surface.

Hope this helps

Thank you Mr Coulet.
The real Mesh I'm working on contains more than 1 million nodes. I can generate water elevation over the mesh by Scilab code accepting linear decrease in water lever upstream to downstream and constant water level on cross nodes belonging to lines perpendicular to centerline (river skeleton). For velocities (U,V) I guess that almost impossible.
Here I stand, I can do no other. I will start simulation with (Ui,Vi) = (0,0) and wait until the correct initial condition is reached.