FEKO warning 121

Altair Forum User

Hi,

 

I requested a far-field pattern where I got this warning:

 

Triangle XXX is too large (surface/lambda^2= 3.45799E-02)

WARNING 121: Coarse segmentation for triangular patches

 

Does this have an impact on the result? any way I can get rid of the warning?

JIF

Hello shashank119,

 

Coarse segmentation of the model can have an effect on the results. The exact amount of the effect depends on the model and what you are calculating. As an example, performing RCS calculation (or far field only calculations in general) is much less sensitive to coarse segmentation due to the integration nature of the far field calculation. Coarse segmentation can have a big impact on nearby near field values and port impedance values. In general, if FEKO gives you a coarse segmentation warning, it is advised to resolve it by meshing finer (possibly only where the segmentation is coarse). Thus, your solutions include meshing the entire model finer or locating the face(s) causing the warning and applying a local mesh refinement. If you use the auto meshing in FEKO, you should not run into these problems (if that is the case here, please provide the model so that we can investigate it further).

 

For extremely coarse segmentation, FEKO will give an error since the answers could then be completely wrong.

Altair Forum User

Hello JIF,

 

This solves my problem. Thanks very much.

 

I'm performing far-field calculations only. So I'm imagining the coarse segmentation is not that much of an issue in terms of accuracy.

 

I used auto meshing in FEKO as my model is electrically large. I tried to initially mesh it myself, but then I had errors where triangle edges weren't of same length (a large triangle for one piece of geometry Vs. a fine mesh triangle of a smaller piece of geometry which were unioned together for example) so I let FEKO do the meshing for me. Unfortunately, due to the nature of my work, I cannot share the model with you. 

 

In my spare time, I can replicate this with a different model and send it through (maybe in 2-3 weeks). 

 

Thanks for your response. Much appreciated.

 

Regards.

JIF

I understand, many of our customers can't share the models. I would really appreciate it if you could replicate the problem in a model that you can share. We are aware that the mesher does not always adhere to the meshing requirements and we are working with them to find a solution to the problem. It has thus happened on occasion that a few elements would result in a warning, particularly if the coarse auto meshing has been used.

Altair Forum User

Hello JIF,

 

Will definitely do my best to replicate and provide you a model as early as possible.

 

Thank you.