[enigma6666]

Hi

Can anybody specify how to define different material for two objects in fluent?

Thanks

[stuart23]

You must have the two objects meshed as separate volumes in your input mesh. You could potentially mesh each volume separately and append one mesh to another (or use Tmerge). Fluent will read your .msh(s) and recognize that there are two interior zones. Then in "Cell Zone Conditions" you can specify the material for the two objects. Hope this helps.

Stu

[Amir]

Originally Posted by stuart23

You must have the two objects meshed as separate volumes in your input mesh. You could potentially mesh each volume separately and append one mesh to another (or use Tmerge). Fluent will read your .msh(s) and recognize that there are two interior zones. Then in "Cell Zone Conditions" you can specify the material for the two objects. Hope this helps.

Stu

Dear stuart,
It is not necessary to have separate zones in input mesh; it's also possible to do that in FLUENT by marking and separating desired volumes.

[stuart23]

Amir, that is true, but not very easy to accomplish if the model is complicated. I would still suggest keeping the volumes separate from the CAD geometry all the way through the mesher and into the solver.

[Amir]

Originally Posted by stuart23

Amir, that is true, but not very easy to accomplish if the model is complicated. I would still suggest keeping the volumes separate from the CAD geometry all the way through the mesher and into the solver.

I don't agree with you! this procedure is applicable for any geometry because you can mark your desired zones in different ways; e.g. for worst cases, you can use temporary memory for iso-value marking.
I prefer this method because you shouldn't concern about grid generation in complicated cases; i.e., using mapped typologies is more probable by this method; furthermore, this method is also useful for omitting threads. For example, try to mesh a pipe with a circular hole on its boundary with mapped topology, which procedure do you think is more efficient?

[stuart23]

Hi Amir,

I'm afraid I do not quite follow what you mean. I understand how to mark enclosed elements using an inequality (isosurface), however I do not understand what value you could use to differentiate between the seperate volumes.

Please clarify this techinque, I am always eager to learn more and more about how different people do things!

Stu

[Amir]

Originally Posted by stuart23

Hi Amir,

I'm afraid I do not quite follow what you mean. I understand how to mark enclosed elements using an inequality (isosurface), however I do not understand what value you could use to differentiate between the seperate volumes.

Please clarify this techinque, I am always eager to learn more and more about how different people do things!

Stu

Hi,

In this method, you need to know where your desired zones are located mathematically; then you need a UDF with an on-demand macro which assign an arbitrary value to your temporary memory (UDM) and then, you can mark and separate these zones via iso-values. As you noticed, regardless of complexities of zones, you can implement this method if you can interpret these zones as mathematical relations; i.e., f(x,y,z)=0 (as its boundary); it's obvious that you can use couple of these relations simultaneously in one UDF. The other restriction is that the boundaries generated are not very exact because it's separated from elements which is expected.
I know that this method is not fully general, maybe in cases in which there are some zones that we can't find exact math relations, but I find it very helpful in many cases which reduces mesh elements and improves its quality considerably.

hope that helps you

[stuart23]

Ahhhhh, I hadn't thought of that. It is a very clever way to separate volume regions. I can see it being difficult for complex geometry though (i.e. freeform surfaces and such)

Stu