Disperion calculation


Dear Lumerical Team,
I have an query related dispersion analysis,
The dispersion calculated in the KB model (or similar Dispersion calculations), sir the dispersion is the total dispersion (means material + waveguide) or waveguide dispersion?
kindly let me
with many many thanks


Dear @Nanda

This is a great question.

There are few things to notice here:

  • If the material index is fixed by either a) setting material index to object defined electric and choosing a fixed value or b) if the material fit returns the same index for all the wavelength .i.e. it is an straight line as a function of wavelength, the observed dispersion will be related to only waveguide (or mode profile). For waveguides with subwavelength cross section, you will observe dispersion due to evanescent field.

  • If material index varies with wavelength, the observed dispersion will be due to material + waveguide. This means that even if you choose to have a waveguide with a very large cross section, you would still see dispersion.

In this example, we have used silica and silicon as waveguide material. Checking the material fits, silica uses a straight line to fit the material, however, index of silicon varies as a function of wavelength:

This means that the observed dispersion is due to material + mode profile.

I hope this was helpful.


Dear Sir,
Thank you very much for your reply,
even I have few quires,
In the above example fit tolerance is 0.1 for silicon and 0.01 for silica, the fitting value effects dispersion? means total dispersion changing into waveguide dispersion.
For one more example from KB base
here the calculated dispersion is total dispersion, here the material is corning glass and in this case the material was used basing on the Sellmeier values and not fitted with option “fit with multi-coefficient model”. But the dispersion is total dispersion mentioned in KB base. Kindly let me.
My mine point is if we kept fit tolerance 0.1 for core and clad then the calculated dispersion is total dispersion or not?

kindly let me
with advance thanks to lumerical team


Dear @Nanda

If the fitted line has changed with fit tolerance, then yes, fitting values affect transmission.

If the material index varies with wavelength, calculated dispersion will be total dispersion either if fit with multi-coefficient model is enabled or not. Only for cases where material fit is an straight line or material has a fixed index, dispersion will be only waveguide part.

If you do not fit material with multi-coefficient mode, but material index still varies with wavelength, observed dispersion will be total dispersion.

If these are confusing, a good technique to distinguish total dispersion from waveguide one is to perform simulation with fixed index value. If results match, dispersion will be only due to waveguide. If not, it would be total dispersion

I hope I could answer your question.


Dear Bkhanaliloo

Thank you for your valuable information.
I am creating my own file and I will send through lumerical support kindly check my file.

with many many thanks