Learning Forum

[vagadiya]

While doing HFSS simulations, In solution data I can see convergence data for all the number of passes, but how can I correlate this with error at every frequency point in my S-parameter plot. I want error bars in my S-parameter plot. How can I do this?

[Praneeth]

Hi,

Thank you for reaching out to the Ansys learning forum.

The adaptive meshing will only be done at single or a set of frequencies depending on the type of sollution frequency. Hence, the change in S-parameters between the passes will only be applicable to those frequencies. Once the adaptive meshing converges, then the frequency response will be determined in the band of the frequency sweep.

Best regards,
Praneeth.

[vagadiya]

I use 5 GHz as an adaptive frequency, and the frequency range is 2.5 GHz to 7.5 GHz. I set delta S = 0.02, but I also want the standard deviation in S-parameter for every frequency point in my frequency range. How can I do that? Because physical quantities without standard deviation are not valid.

[Praneeth]

I did not understand your query. Please let us know what physical quantities you are referring to help us serve you better. Are you looking to do any statistical analysis with respect to any of the design variables in your simulation model? If possible please share the reference for a better understanding.

[vagadiya]

Here I have this plot of Sdd11 and Sdd22 v/s frequency. From the graph at 1 GHz value of Sdd22 is approx. 0.1 but this 0.1 is just a value it does not have any standard deviation(error) in it. Whenever we write any physical quantity we write for example 0.1 ± 0.05. But here in this plot I have only 0.1 so how can I extract this error values in each frequency point??

[Praneeth]

What error you are looking at and what is the source for it?

Are you referring to DeltaS as error from the convergence table? If so, as stated before, in the convergence table, you will see the difference between S-parameters at the solution frequency between two concecutive passes of adaptive meshing. It does not denote the error in calculation.

Best regards,
Praneeth.