[ahmed ramadan]

My case is very close to this case
l m going to simulate oscillating flow in circular pipe L=5D
At Re_tau=1440 with pressure gradient not constant value
i used periodic boundary conditions so the fluent doesnt allow to input UDF for pressure gradient if periodic just specific one
can any one know any way to go across this case
thank u all

[sbaffini]

Dear Ahmed,

You just need an UDF with a momentum source term dependent on time. This just represents the opposite of the pressure gradient you would tipically use.

Find attached a trivial example (i have not tried compilation, but it should work with most Fluent releases)

[ahmed ramadan]

First: thank you for your time
But in the periodic boundary there is no space to compile UDF i cant inter one
And my pressure gradient dp/dz=a.cos (wt) and a varies from -4&4
I think that is little bit different
thank you again

[sbaffini]

Notice that you don't need to use it under the periodic panel. Just use the source term in a classical sense. This is the way the pressure gradient is fixed by Fluent itself (it's just a source term)

[ahmed ramadan]

Sorry i cant got it !
Would you give some details
Thank you

[sbaffini]

Well:

1) Setup periodicity as usual but leave the pressure gradient as 0

2) Add a momentum source term using the supplied udf

In order for this to work as expected you have to modify the udf so that the source term has the same direction of your intended pressure gradient, the same magnitude and an opposite sign.

[ahmed ramadan]

Dear Sbaffini if you are using periodic boundary condition there are no place at fluent to add momentum source term and there is no place available to add the udf for pressure gradient .... just specified number

[sbaffini]

Please, read the UDF manual on how to add a momentum source term in Fluent. If you know how to do that, you can certainly use my udf above for thst purpose. Once you do that, addong periodicity with 0 pressure gradient is trivial. I would help with pictures but i am on the cell phone now

[ahmed ramadan]

How are you dear Sbaffini
I tried your advice it works well at the steady part but at the unsteady one there is some thing wrong
And i have a question when add the source term to the axial momentum it means to add this source term to all nodes on the internal zone not for the periodic nodes which is boundaries
Is that wright.

[sbaffini]

Dear Ahmed, what do you mean by something wrong in the unsteady part?

You just need to add this as classical source term
(Fluent knows how to do that) under cell zone conditions.

I am on the phone again but i will provide some screenshots as soon as possible

[ahmed ramadan]

I add the udf to the cell zone condition to the axial momentum but i m asking about that to understand it physically

[ahmed ramadan]

It means fluent will read this udf at every node of the internal node
And will read 0 for pressure gradient at the the periodic boundaries

[sbaffini]

Well, in finite volume, the source term is applied on a cell by cell basis, just like the pressure gradient. This means that it is simply applied to all the cells.

[ahmed ramadan]

Ok
I will read this part well and complete this discution
And thank you very much for your time

[sbaffini]

I mean, there is no node implied, just cells.

However, for the record, in a finite difference setting you would have used the source term for each of the nodes you are actually solving for.