CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FLUENT (
-   -   initialize 3D domain with 2D results (

ivanbuz August 28, 2009 03:21

initialize 3D domain with 2D results
I have a 3D domain which is created by simply repeating the 2D domain in the z direction. if the 3D domain is initialized by giving a uniform value to the whole domain, it is very time-consuming to develop the flow.

here comes the question: can I initialize the 3D domain using the 2D result which is easy to obtain?

maybe many of you have had the same question. please throw me some hints, thanks!

coglione August 28, 2009 07:12

Hello Ivan,

i had a similar problem and writing an interpolation file from 2-d and reading this back into 3-d did the job.


ivanbuz August 28, 2009 13:09


Great! I need more details to really know how to do it. how to write the interpolation file? do you expand the 2D flow and make it contain 3D flow field info in the interpolation file?

coglione August 31, 2009 03:05

No need to expand to 3-D, have a look at the user guide for writing and reading interpolation files.


ivanbuz September 1, 2009 16:20

Hi, Max,

I did what you said (writing an interpolation file from 2-d and reading this back into 3-d), but I got a error message saying "the interpolation file has wrong dimensions". any way to get around this?

coglione September 3, 2009 04:13

Hello Ivan,

i have to apologize for my last post which was incorrect. Actually you have to edit the 2-D interpolation file by hand:

1) At the second line change 2 to 3 to tell Fluent it is for 3-D now
2) You need to add a list of z-coordinates after the list of y-coordinates. This z-list should have the length of your data-points (number at line 3 = cells in 2-D) and the value of each point may be an arbitrary number but should be constant.

Have a look at the user guide where the syntax of interpolation files is well explained.


ivanbuz September 3, 2009 18:19


You are very helpful, thank you very much for giving the details!

I tried another way which might be a little simpler, and it seems to work well: use a smaller 3D interpolation file to initialize a bigger 3D domain. Note that the two grids are different only in the z direction -- the bigger grid has a bigger z. I have NOT experimented on grids whose geometries are different in all three directions.

All times are GMT -4. The time now is 08:43.