CST2013: Dielectric / Magnetic Dispersion Fit

Modeling: MaterialsNew/EditNew MaterialConductivityDispersion[Dispersion List
Edit Object Properties (navigation tree: Materials:material1PropertiesDispersionDispersion List)

This dialog offers the possibility of defining a specific material dispersion curve, that is automatically fitted to the specified material model. The following description applies both to the dielectric and magnetic dispersion fit, as the features, control and material setup is exactly the same.

Eps (Mue)

Fitting scheme: Several different dielectric (magnetic) dispersion models can be chosen. The models are: Conductivity, General 1st order, General 2nd order and General Nth order. For detailed information about these schemes see the Material Overview.

Frequency/Value'/Value'' (Frequency/x'/x''/y'/y''/z'/z'') / Weight: In this list box, the specific eps' and eps'' (mue' and mue'') can be defined by setting several values at different frequency points. Moreover to each frequency a weight (value greater than/equal to 0.0) is assigned in order to direct the interpolation algorithm and to enforce a reduced error in correspondence of the given frequency point.

In case a conductivity model is given for the dielectric dispersion the relationship between the eps' and eps'' value required in the list box and the conductivity value may be found in Material Overview (HF) - Conductivity Models.

An important observation is that in correspondence of General Nth order fitting, due to specific algorithm requirements, the weight behaves as a switch between considered and ignored samples. A weight less then 1.0 forces the interpolation algorithm to ignore the given sample, whereas a weight greater/equal 1.0 to consider the sample itself.

Max order: This control is available only in case of General Nth order fitting. It enables the user to specify the maximum allowed model order (total number of real and complex poles each counted with the corresponding multiplicity). The search of the best fitting model is then stopped to the given maximum order even if the error convergence criterion is not fulfilled.

It should always taken into account that increasing the model order may improve the quality of the fitting but at the same time the simulation complexity in terms of memory and computational time.

Used order: Displays the effective model order provided by the interpolation algorithm. The model order corresponds to the total number of real and complex poles each counted with the corresponding multiplicity (in formulas, Order = num Real Pole + 2 x num Complex Poles).

Error limit: This control is available only in case of General Nth order fitting. It enables the user to specify the error stopping criterion during the search of the best fitting model. The error should be interpreted as the maximum relative error computed in correspondence of each given frequency point between the complex epsilon (mue) value and the fitting curve. A useful interpretation of the Error limit is as "error" or "inaccuracy" during measurement of the real material properties.

Error: Displays the obtained fitting error. The error should be interpreted as the maximum relative error computed in correspondence of each given frequency point between the complex epsilon (mue) value and the fitting curve.

Details...: Opens a dialog displaying information about the fitted model, in terms of position of  zeroes and poles in the complex plane.

The representation is subdivided in first and second order contributions. The first order contribution corresponds to a real pole, whose frequency is computed and shown.

The second order contribution may derive from a couple of real poles or from a couple of complex conjugate poles. In the former case the frequency of the poles is computed. In the latter case the resonance frequency of the pole together with the quality factor (Q) is shown.

The description of these parameters and the mathematical formulas that relate the first and second order polynomial coefficients to the physical interpretation of poles, zeroes and resonance frequencies may be found in Material Overview (HF).

Use data in frequency range: This control, available in case of General Nth order fitting, allows the fitting algorithm to use only the frequency data points that lie within the "frequency range settings" defined by the user (see also Frequency Range Settings). Activating this check box enables an accurate data fitting of the material resonances which occur in the simulation bandwidth of interest using possibly a reduced number of poles and zeroes with respect to the complete data fitting. And this, in turn, translates into benefits for the simulation complexity in terms of memory and computational time.

Load File...: Opens a file selector dialog to specify a ASCII file containing point coordinates.

The file is formatted in lines which specify the frequency points. In case of isotropic "Normal" material, each line has four entries specifying respectively frequency, value of real and imaginary part of eps, i.e. eps' and eps'', (mue, i.e. mue' and mue'') and weight. The weight field may be skipped and is assumed by default equal to 1.0.

In case of "Anisotropic" material, each line has eight entries specifying respectively frequency, real and imaginary value for X/Y/Z eps (mue) components and weight. Also in this case the weight field may be skipped and is assumed by default equal to 1.0.

Moreover it is possible to set a default frequency unit scale which is applied to all loaded data samples.

Delete: Press this button to delete all selected rows in the dispersion fit list box.

Clear List: Press this button to delete all entries of the dispersion fit list box.

OK

Accepts the input and closes the dialog.

Cancel

Closes this dialog box without performing any further action.

Apply

To apply the settings and show the fitted curve.

Help

Shows this help text.