What's the largest number of tests you solved with HFSS before having memory problem?

[Deepa.r]

hfss spice time steps per rise time

What is the maximum dimension of the structure that can be solved in HFSS v.8.0.25?
Simulating a 80mmx80mm planar structure in hfss doesnt yield the required results.

What would be the suggested no. of tetrahedron for meshing?
I have tried upto 40000 tetrahedron without any positives..



Plz reply

 

[peleda]

hfss symmetry

Zotti's Question -
The problem is with the lambda refinement requirement. That is my guess. I had a similar case, which I am still working on. I set the lambda to 10%. And I was running it on a cluster. It did not go through. When I left it @ 25%. Everything was fine. Decrease your level of accuracy, and see what happens. Remove any manual mesh refinement, and see what happens.

Then slowly refine the simulation. Sometimes you do not need to over refine your results. If after the first refinement, you do not see a change in results - do not bother going far.

Added after 6 minutes:

Deepa,
I have been able to simulate a structure that is about 10m by 2m by 11m. In fact, it was an aeroplane. At about 60MHz. This was in HFSS10. There should not be much difference. Well kind off - they changed the type of meshing.

 

[hgc1979]

numerical method used in hfss

peleda,
i agree with you

 

[aw]

hfss matrix size versus tetrahedrons

hi peleda

60 MHz for an aeroplane is not paricularly impressive - it's a pretty low frequency - the structure may be large but no longer electrically large. There are better tools on the market for the simulation of large structures and at realistic frequencies.

aw

 

[peleda]

insufficient memory error in hfss

"aw",
out of curiosity, what are the better tools on the market for this. I would seriously like to compare performance.