[ahmyaser]

Hi,

I am trying to set an inlet boundary condition for a simulation similar to wind tunnel testing.

The inlet boundary condition for the velocity should be a table or list (that I have already generated using some numerical models)

The table has the velocities as a matrix that has a number of rows equal to the inlet cells or faces, and a number of columns equals to the number of time steps for the LES simulation.

I will be using the pisoFoam model and I wonder what type of boundary condition to map these list that varies with location and time step.

I tried to use nonuniform List<vector>, but it only varies with location and doesn't vary with time

Also, the timevaryingmappedfixedvalue varies with time and doesn't vary with location

Can anyone help me with this please?

[piu58]

> and doesn't vary with time

There may be more advanced techniques, but there is also an easy way: Stop the simulation after each time step, use a script to change the b.c. and start the next time step.

[ahmyaser]

Originally Posted by piu58

> and doesn't vary with time

There may be more advanced techniques, but there is also an easy way: Stop the simulation after each time step, use a script to change the b.c. and start the next time step.

Thank you very much piu58, Can you give me a quick guidance on how to to change the b.c each time step ???

Thanks again

[piu58]

I don't konw in which form your b.c. exists. The easiest way is to prepare pattern files for all the b.c. objects in your 0 directory in the form

internalField   nonuniform List<scalar> 
32336
(
#0
#1
#2
:
)

You may replace the #0 and so on by a sed script which you generate from your list:

's/#0/myBoundaryCondition_0/'
's/#1/myBoundaryCondition_1/'
:

[ahmyaser]

Originally Posted by piu58

I don't konw in which form your b.c. exists. The easiest way is to prepare pattern files for all the b.c. objects in your 0 directory in the form

internalField   nonuniform List<scalar> 
32336
(
#0
#1
#2
:
)

You may replace the #0 and so on by a sed script which you generate from your list:

's/#0/myBoundaryCondition_0/'
's/#1/myBoundaryCondition_1/'
:

Thank you very much, piu58. for your great help and sorry for asking so many questions as I am new somehow to openFoam. I just wonder how can I let the solver know that this list is assigned to that time step? in other words, how do I link each time step to a certain list of inflow?

Your help is so much appreciated !!

[piu58]

The solver read what it is in the <timestep> directory. At the beginning this is the 0 folder. You have to manipulate / replace the files in the directory of the last timestep.
Of course you have to set the startFrom entry in the controlDict file to latestTime.

[ahmyaser]

Originally Posted by piu58

The solver read what it is in the <timestep> directory. At the beginning this is the 0 folder. You have to manipulate / replace the files in the directory of the last timestep.
Of course you have to set the startFrom entry in the controlDict file to latestTime.

That is a great idea!! thanks a lot. you are a lifesaver

[ahmyaser]

Originally Posted by piu58

The solver read what it is in the <timestep> directory. At the beginning this is the 0 folder. You have to manipulate / replace the files in the directory of the last timestep.
Of course you have to set the startFrom entry in the controlDict file to latestTime.

just a last (hopefully) question to make this replacement for each time step. where should I do a script for that? is it at the controlDict derctory?

[piu58]

You need to let the simulation run by a shell script in the form
loop:
- prepare the b.c. with sed
- start pimpleFoam or what solver you want to have (needs to stop after one time step)
endloop

[ahmyaser]

Originally Posted by piu58

You need to let the simulation run by a shell script in the form
loop:
- prepare the b.c. with sed
- start pimpleFoam or what solver you want to have (needs to stop after one time step)
endloop

Thanks, can you let me know what is (sed) ?