Learning Forum

[Conor Leighton]

Hi,
I have an array of components in a fluid domain (order of hundreds), each of which has cylindrical material properties.
I have created local coordinates in spaceclaim that would be the origin for the material properties. This workflow works in mechanical.
However, I understand that fluent works entirely on the global coordinate system - and as such, each component would require the material property axis to referenced according to 

Is there a way of exporting the local coordinates from spaceclaim into something fluent can understand?
Is there a way of 'looping' through each components applying the material properties appropriately.

I have a horrible feeling I may need to start getting my head around UDFs and python . . . but if anyone could give some examples as a head start I'd very much appreciate it!

[Rob]

No, Fluent can also work on local coordinates. It depends on what these components are/do. 

[Conor Leighton]

 

OK, that sounds promising, cheers Rob.

The model has an array of rods, each of which have self heating and cylindrical anisotropic thermal conductivity centred about their axis.

The ‘desire’ for local coordinates is to have one per cylinder, which I can then use to assign the axis of the cyl-orthrotopic material properties to – though I’m open to other methods if they are preferable?

There’re no sliding meshes or anything that fancy – just a load of cylinders getting hot. The fluid domain is non-repeating so I can’t just model one unfortunately.

It was relatively trivial to attach the coordinates in spaceclaim – and mechanical can use them in exactly this way. I’m struggling to find the equivalent method in fluent.

Any and all help greatly appreciated.

 

[Rob]

I think you set the material via 

So it's material dependent. However, once you've set one material up the TUI comes in to play. Click into the Fluent Console, and type

/define/materials  

Then follow the prompts. This will give you the command syntax. It's then a case of combining the commands with the coordinates & material labels from SpaceClaim etc to build a journal file. Sounds complex, it's very simple but you need to exercise care. 

There's also a solid zone option - Material Orientation - but I don't that's what you need here. 

 

 

[Conor Leighton]

 

OK, I set the material properties as above – I have defined one material:



I have then tried to assign that material to each component using a journal file, stating the axis of rotation about which I would like the material to orient. The script has indeed applied the material and altered the coordinates for each cell:




However, the results clearly show that this hasn’t worked – it would appear that every cylinder has the cylindrical axis for thermal conductivity centred about the global origin, rather than around the coordinates specified in the solid zone rotation-axis origin.

Not quite sure what it is I am doing wrong – the only thing I can think of is that I have to create a new, independent material for every single cylinder and then alter the coordinates in the cyl-orthotropic panel – but this doesn’t really feel sensible?

 

[Rob]

I possibly wasn't very clear. You need a unique material for each zone. It's a fairly new feature, so you may want to test on a smaller model. 

[Conor Leighton]

Hi Rob,

Thanks for your continued support - much appreciated!

So what I have been doing is defining the material once, then looping through the cell zones assigning the material, and then (wongly) using the cell rotational zone coordinates, assuming that that aligned the cyl-orthotropic axis to the cell zone.

What I need to do is loop through defining the materials with the same coordinates I previously used, and then another loop only assigning the material properties but leaving the cell rotational zone coordinates alone.

OK, cool, thanks Rob!

[Rob]

Probably - it's Friday....  So, lots of materials each with the position coordinates. Then assign each cell zone it's (correct) material. 

Let me know how you get. 

[Conor Leighton]

It works!

Well, at least that issue works. Results show the expected behaviour.

In essence for further readers: one material per component, with the cyl-orthotropic coordinates in the material properties. Even if all the components are the same - the material property defines the coordinate system for the assembly.

Cheers Rob, much appreciated!