[IrieConqueror]

hey guys,

i have a tank within a fluid. i got a few nozzles inside as an inlet. there is a tube at the wall as an outlet. i get smooth convergence for steady state. now i want to examinate the mixing time, when i put tracers into the inlet for 10 seconds. i tried transient simulation, with an initialfile of the converged solution of the before made simulation.
any suggestions how to add the tracers that passed the outlet to the inlet? i tried:

"inlet="step((10[s]-Time)/1[s])+areaAve(Tracer)@outlet

but therefore i got tracer imbalances between 1.5 % and - 3 %.

thanks in advance

irie

[ghorrocks]

What do you mean by " how to add the tracers that passed the outlet to the inlet?" Are you asking how much tracer has gone through the outlet?

And what do you mean by tracer imbalances?

[IrieConqueror]

thanks for your reply and sorry for my english. imagine that the system should be a circulation and the tracers that passed the outlet need to be added to the inlet flow.
i'm not quite sure if i get it right with the imbalance. in the attachment you can see the imbalance for the transport equation of the tracers. do this imbalance has something to do with my self created condition "add outlet-tracers to the inlet" or is it just because of "numerical problems"? how would you realize a circulation or do you think it is right to do so?
update: i reduced the timestep, increased the loops and was able to reduce the imbalance a little but the course of the graph still looks the same.

[ghorrocks]

This imbalance on the tracer variable is simply a convergence thing. The tighter you converge the solution the smaller the imbalance will get. So just run more iterations and it will get smaller. Also, remember the key things which help convergence and will make it easier to get the imbalances down: improve mesh quality, double precision numerics, better initial condition, smaller time step size (if transient).

[IrieConqueror]

thanks again! i can comprehend what you said in case of steady state simulation. i want to examine the mixing time and the progress and most of the mixing is done after 300s, why i "can't wait for convergence" in case of imbalance. i got results after every 5 seconds to obtain the concentration.. the results already converged in my initial file (steady state simulation) which i used in the transient simulation. i'm confused

[Opaque]

Careful with the interpretation of imbalances.

Conservation imbalances by definition must ALWAYS be 0, either for transient or steady simulations. It is not zero because the solution is not fully converged in the conservation sense, only in the user-defined residual sense. It has nothing to do with the inflow/outflow.

Now, since there is no transient term for steady-state simulations, as a user, we can connect the "conservation imbalance sense" with the "inflow - outflow imbalance sense".

For transient simulations, the difference between "conservation imbalance measure" and the "inflow - outflow imbalance measure" is the transient term which may not be zero if the flow is unsteady.

[IrieConqueror]

thanks for that explanation. i think i get closer to understand but still got problems. maybe i have to look closer to the transport equation to fully get it. not that easy.
do you know any good literature?

[ghorrocks]

The CFX documentation explains all these concepts well.