Tank refuelling simulation
 

Hi, I need to simulate a tank refuelling process.

The tank has 2 pipe (an inlet pipe for Adblue chemical and an outlet pipe for air, which leaves the tank as adblue occupy it's space). The position of both pipes is on the side of tank not on top (air outlet pipe is positioned slightly higher than the Inlet pipe). One end of the air outlet pipe is attached to the tank and other one on the side of adblue inlet pipe just above the position where my inlet boundary is.
In practice, during tank refuelling with nozzle when the adblue level reached to the air outlet pipe in tank, it also get filled with it and subsequently the inlet pipe also. At this time the nozzle shuts off. But due to compressed air in the tank few drops of adblue spit back through inlet pipe to the person filling the tank.

To simulate this phenomenon, I choose Multiphase flow, k-epsilon turbulence model with the initial condition of 90% volume of tank is full.
But the main problem is how to shut off the inlet flow during simulation when both inlet and outlet pipes are full as it happens practically and continue to run the simulation for another 1 s to simulate the spit back behaviour of drops.

Do I need a user defined function for this? I have no idea how to write a user defined function for this.

Hi Kamal

your problem is fairly challenging (atleast for a fledgling like me). Do you know at what time step when the tanks and inlet and outlet pipes are full?

If yes, then you can specify the field function that will convert the inlet boundary to wall.

This can be done using a macro very easily too. Record a macro with Single time step. After one time step is over, convert the inlet boundary to wall boundary, and again hit Run on the ccm+ interface.
Stop the macro and change the number of desired time step in the java code.

This will do the job

Hope this was helpful.

Hi Gajendra,

Thanks for your answer. Actually I don't know the exact time at which my tank and pipes get full. During experiments, there are lots of factors which affect the Mass flow rate, so I can't depend on that data. Does it mean that I have to do it with hit and trial method?
To change the inlet boumdary to wall, do I need to stop the simulation and keeping the macro running?

Hi

Had i been in your place, I would have initially simplified my problem by assuming constant rate of volume flowing through the pipe (assuming uniform flow through the pipe) to the tank and done a hand calculation of the time at which the tank would have gotten full and used the macro

Regarding macro, you just have to play the macro once, which will run your simulation and then in the specified time-step will convert the boundary to wall by itself and then continue the iterations further.

Thank You

Thanks, I will try to simulate it as you suggested.