Learning Forum

[Luca Sacchi]

Hi 

I am simulating with FDTD a library of Pillars of SiO2 on a substrate of SiO2 and everything is coated with a thin layer of TiO2. I am having difficulties to set the right parameters for the distances of the Transmission monitor, PML boundary distances from the structure, the simulation time, autoshutoff value and mesh size. 

The problems arise because, the thin coating makes the structure resonate a lot and it takes a lot of time to finish the simulation because it has reached the autoshutoff value. Moreover to resolve the thin coating I have to use a fine mesh. I have to simulate around 40.000 structures and I am trying to find the optimal values to make the simulation region as small as possible and the simulation as short as possible. 

Which values for "simulation time", "autoshutoff" should I use? Currently I am using 10000 fs, 1e-5. 

What is the optimal distance of the trasmission monitor from the top of the pillar? I have to record the field to then do farfieldexact, and I want to record only the far field and not any possible near evanescent field.

How far from the transmission monitor and/or top of the pillar should the top PML boundary be? 

How far from the reflession monitor and/or bottom of the pillar should the bottom PML boundary be? 

What is the minimum number of Yee cells that should be present in the coating in order to resolve it? At the moment I am using a mesh of 8nm defined around the pillar and coating. 

The xy size of the FDTD is 450nm. The Pillar is 2um tall and I am simulating broadband 400nm-700nm. 

Best

Luca

[Niki Papachristou]

Hi Luca,

Thank you for posting on our forum! Your questions are really good and these are very common when the users are starting up with their simulations! Unfortunately, I can't give you the excact numbers for all of your questions, simply beacause they depend on several parameters that I can't predict. This also depends on the time and memory requirements of your simulations. Your questions will be answered simply by consistently trying different cases and ending up to meaningful results. One good starting point is the Convergence testing process for FDTD simulations article.

I hope that was helpful, let me know if I can be of further assistance.

Please note that this is the Optics forum and not the photonics where lumerical related question are posted. For your next lumerical question please post at: Ansys photonics forum

Kind Regards,

Niki