Hello,
In your listing file, you can isolate a few lines for every printing in the listing file:
DIFFUSION OF WN
GRACJG (BIEF) : 44 ITERATIONS, RELATIVE PRECISION: 0.9348234E-04
>>> That concerns the diffusion step of velocity with conjugate gradient solver "GRACJG" (SOLVER FOR DIFFUSION OF VELOCITIES = 1, default value), MAXIMUM NUMBER OF ITERATIONS FOR DIFFUSION OF VELOCITIES = 400 and ACCURACY FOR DIFFUSION OF VELOCITIES = 1.E-4. You can read that 44 iterations are needed for this step, which is OK (not too much but enough).
PROPAGATION AND DIFFUSION WITH WAVE EQUATION
GMRES (BIEF) : 1 ITERATIONS, RELATIVE PRECISION: 0.9308448E-04
>>> That concerns the wave equation step of velocity with GMRES solver (SOLVER FOR PROPAGATION = 7, default value), MAXIMUM NUMBER OF ITERATIONS FOR PROPAGATION = 400 and ACCURACY FOR PROPAGATION = 1.E-4. You can read that 1 iterations is needed for this step, which may be not enough, in particular if the asked accuracy is quite big. In your case, I would recommend you to decrease the value of ACCURACY FOR PROPAGATION to 1.E-6, even default value 1.E-8 if the number of iterations is not too big (a good target is around 10 to 30).
DYNAMIC PRESSURE STAGE
GMRES (BIEF) : 62 ITERATIONS, RELATIVE PRECISION: 0.9286598E-03
>>> That concerns the Poisson Equation step for Pressure with GMRES solver (SOLVER FOR PROPAGATION = 7, default value), MAXIMUM NUMBER OF ITERATIONS FOR PROPAGATION = 200 and ACCURACY FOR PROPAGATION = 1.E-3. You can read that 62 iterations is needed for this step. Perhaps you can decrease the time step (divide by 2 e.g.) or you can try another solver SOLVER FOR PROPAGATION = 1.
ADVECTION-DIFFUSION OF K-EPSILON OR OMEGA STEP
DIFFUSION OF AKN
GRACJG (BIEF) : 2 ITERATIONS, RELATIVE PRECISION: 0.3134408E-04
DIFFUSION OF EPN
GRACJG (BIEF) : 2 ITERATIONS, RELATIVE PRECISION: 0.1156087E-04
>>> That concerns the diffusion step for the 2 turbulence variables k and epsilon with conjugate gradient solver "GRACJG" (SOLVER FOR DIFFUSION OF K-EPSILON = 1, default value), MAXIMUM NUMBER OF ITERATIONS FOR DIFFUSION OF K-EPSILON = 400 and ACCURACY FOR DIFFUSION OF K-EPSILON = 1.E-4. You can read that 2 iterations are needed for this step, which may not be enough. In your case, I would recommend you to decrease the value of ACCURACY FOR PROPAGATION to 1.E-6, even default value 1.E-8 if possible without too many iterations (a good target is around 10 to 30).
ADVECTION-DIFFUSION OF TRACERS
DIFFUSION OF TRN1
GRACJG (BIEF) : 0 ITERATIONS, RELATIVE PRECISION: 0.7289564E-04
>>> That concerns the diffusion step for the the tracer with conjugate gradient solver "GRACJG" (SOLVER FOR DIFFUSION OF TRACERS = 1, default value), MAXIMUM NUMBER OF ITERATIONS FOR DIFFUSION OF TRACERS = 400 and ACCURACY FOR DIFFUSION OF TRACERS = 1.E-4. You can read that 0 iteration is needed for this step which is clearly not enough. You do not know if the solving step has converged or not. For tracer, you should try the default value for ACCURACY FOR DIFFUSION OF TRACERS = 1.E-8, even less if only 1 or 2 iterations for this step. If with 1.E-8 you need more than 50-60 iterations, you may try to increase a little bit (e.g. 1.E-7 or 1.E-6) but I would not recommend to ensure tracer conservation.
The messages concerning JACOBI and MURD_LIPS are here for the use of LIPS advection schemes (default advection schemes for TELEMAC-3D), in a first step, you do not need to read them if the relative precision is OK.
Hope this helps,
Chi-Tuan
