Learning Forum

[bartu.yaman]

Hello,

 

I have a mode that was calculated in COMSOL and I want to simulate the evolution of this mode with Lumerical EME. I have the individual electric field components and I imported these fields to Lumerical as a rectilinear dataset, so no problem so far. The problem is that when I import the rectilinear dataset to an EME port, the values of the field changes and this messes up the S-matrix values at the end of the simulation. The electric field values I have go from -10 to 10 (as I import them to Lumerical as a rectilinear dataset) and when I import this field to the EME port I see that the values are now somehow in the order of 10^11 (as I see it by visualising the fields again in the port). 

 

What is going on during the transition from rectilinear dataset to the EME port? 

 

If I have electric field data from another software, how do I make sure EME uses this data properly as if the fields were calculated directly in Lumerical?

 

Please tell me precisely how these are defined, I am unable to use the software because of vague/non-existent definitions. 

[Guilin Sun]

It seems there are two issues:

1: the magnitude

2: the mode profile.

In EME, we normalize the fields to have power of 1Watt. So no worry about its values.

For the mode profile, as you know, Lumerical uses rectilinear  meshing. Mostly the other tool uses finite element meshing. So before importing the fields, please view the dataset and make sure it does not change the field profile after interpolation.

If you only have E-fields, EME will recalculate the H-fields from discrete Maxwell Equations.

So before importing make sure the dataset gives the correct profile.

BTW: in almost all the cases EME should be able to calculate the source mode by its self. You may also try to calculate the modes inside EME to avoid importing.

 

[bartu.yaman]

I am importing the mode beacuse I don't have the structure in Lumerical, I only have a mode that was outsourced. I handle the mesh conversion in Python so I am sure I have the same mode as soon as I import it to Lumerical script workspace, so far there is no issue. The problem is that the field magnitudes change as soon as I load the same field in the script workspace to the EME port. I also somehow get transmission values in the order of 10^4. 

[Guilin Sun]

Please extract the imported source data and recalculate its power. As mentioned previously, EME assumes mode power of 1Watt. Maybe your imported fields have larger than 1 v/m electric fields. However the solver assumes it is 1v/m so yo increase its amoplitude to 1e^11. Usually it is on the order of 1e7. This might be the reason 1e^4.

You can either nromalize the import fields, or normalize the result. eg, S11=S110*Mode_power/Calculated_mode_power.