Learning Forum

[232221001]

Hii,

I'm working on a simulation workflow where I perform a 3D optical simulation in FDTD and want to import the resulting carrier generation rate into the CHARGE solver for electrical analysis.

In FDTD:

• The simulation region is 3D, and

• The CW generation monitor is set up with a Y-span of 1 µm.

However, when I import the resulting .mat file into the CHARGE solver:

• The "generation" object appears as 2D (rather than 3D), my simulation region of CHARGE solver is 2D with Y-axis normal.

• And the Y-span is shown as only 0.2 µm, even though the FDTD region spans 1 µm.

I've also looked at an example file from Lumerical where the generation profile appears as a 3D distribution in CHARGE, so I believe this should be possible.

🔹 Could you please clarify the correct and complete steps to export generation data from a 3D FDTD simulation and import it properly into CHARGE as a 3D profile?

I would really appreciate it if you could:

• Provide a step-by-step guide, especially from a basic and design perspective (i.e., what to set in the monitors, export options, and import settings),

• And describe how to avoid CHARGE treating the imported generation rate as 2D when it should be 3D.

 fig 1. Monitor setting of charge generation in FDTD

 

Fig 2 geometry of generation monitor because of imported data from FDTD 

[Dimitris Polyzos]

Hello, 

I would recommend looking into the vertical photodetector example simulation files which exactly uses the analysis group in FDTD to export the generation in the correct format to a MATLAB file, which then can be imported into a 3D generation object in CHARGE. this approach for importing the generation data into CHARGE. Normally, if a structure is uniform along its length, the 3D optical generation rate from FDTD is averaged along the length, which allows doing 2D CHARGE simulations, significantly reducing simulation time. For your case you should avoid averaging this dimension. Thus, you should choose the "average dimension" to set to "none" in order to omit averaging one dimension of generation rate and produce a 2D profile. 

I hope I helped.

Regards

Dimitris