In CST 2014, Mesh too dense (++ millions)! Can't simulate! Please help me.

[dovod]

Hello,

I am working with CST 2014 and I am trying to simulate a rectangular lens coupled to a rectangular horn antenna with a little angle of flange in E Plane and a wider angle in H plane (close looking to a H-sectoral antenna) but the meshing in the aperture and in the lens is very dense, due to the little angle of flaring in E plane. The thing is that even reducing to the minimum (2 cells per wavelength, 1 cell per max model box edge) the global mesh properties, I get a few millions of cells (about 28 m) and it last for a long time to compute just the mesh coefficients.

It seems to me there should be an easier way to do this. Besides, I need to simulate a parametric sweep of the flare angle so imagine how worried I feel about this...

So, please gentlemen, How do I reduce the dense number of cells in the lens and aperture? I have tried with adaptive mesh but when I click param sweep it looks like it ignores it.
Need your help, this is really important for me to work out. Just tell you that the whole thing is very large (depending on the flare angle, but it exceeds at least 10lambda) so I use T-solver, if it helps.

Thank you all in advance!

 

[gyz182]

Hello,

What frequency are you operating at? For electrically large models, try using the I-solver.

You could also try simulating the horn(if that is not too large) and monitor the fields around it (field source monitor).Then import the field source into a new simulation with the lens with it. That should save you some mesh in the models.

 

[kela3kela]

I have the same problem. Im working around 500 THz wt a nm dimensions..
"Unable to generate MESH" and "mesh is too dense or calculation domain may have zero volume".