[TfG]

Hi,

I am simulating the cooling process of an oil tank submited to ambient air and I am having some troubles. First of all, i have increased the number of iterations per time step to 170, and each time step is 2s. With a mesh of 113000 elements (tank is 18m high and 23 diamater, too many elements?), having the first node 0.5 mm from the wall, I am not sure with criteria should I set for convergence in each step. The continiuty scaled residuals are constant slightly above 1e-3. For x and y scaled residuals I set 1e-3 for convergence criteria.

I have been simulating 200s and if I force to perform 170 iterations in each time step I get similar result with 3 meshes (17000, 41000 and 113000 elements). But if I set the convergence criteria of above (I dont check for continiuty because is always constant...?) I get different results: max velocity is half (order is around 1e-5) compared with no convergence checking simulations.

I have simulated checking convergence (otherwise it takes a hell of time) for the mesh of 113000 around 8000s and I have realized that the integral of mass imbalance starts to drop around 3300s. What could be the reason dor this? the residuals of continiuty rise slightly in that instant.

I would like to know the meaning of mass flow rate for "default-interior" too if anyone knows.

I post some pictures of scaled residuals, integral of mass imbalance, max velocity and heat loss (look that also for 3300s the behaviour changes dramatically for velocity and slightly for heat loss: so I guess something goes wrong there).

Other data: model:laminar, boussinesq approx, power law for momentum and energy, body force wighted for pressure, PISO for coupling, URF 0.3 pressure and 0.7 momentum.

I would appreciate your comments and suggestions because I don't know how to progress with this.