July 11, 2014 at 6:53 am #12327PING HEParticipant
I am currently running a Si waveguide grating simulation. I am facing some problems and appreciate any hints from you.
1. The layout size: I am trying to put the whole design into a window as small as possible. However, it seems to be that for different size of the layout, the simulation results are different (I am using the same Mash size). Sometimes it could be a big difference. And it also seems the observation plane can not be too close to the boundary. Otherwise, the reading could be completely misleading. Is this true? And is there anything to do with the boundary condition?
2. About the mash size: I am currently using the default mash size which is approx. 45nm. However, when choosing the smaller or larger mash size, readings such as the angle of diffraction (from a Si grating) could be varied. I am wondering, in general, is mash size the smaller the better/more precise? Or there will be big accumulated error when choosing smaller mash size? If so, dose it mean for each of design, the mash size needs to be tested and chosen within a certain range which may lead to a more precise result?
July 14, 2014 at 7:11 pm #12370Damian MarekParticipant
Mesh size definitely plays an important role, for determining the maximum frequency that can be simulated and also how well a 3D object can be represented, like a sphere.
Even at PML boundary conditions there will be reflections, albeit smaller, so you need to make sure that your simulation domain is large enough to keep reflections at a minimum. Keeping Observation areas away from boundaries is generally a good idea because of these reflections.
From my understanding as long as the time step is meets the Space-Time Stability conditions, a smaller spatial mesh will result in higher accuracy.
July 23, 2014 at 3:59 pm #12846Aurelien DuvalParticipant
you can also modify the properties of the PML layers to reduce further the reflections. To do this, go to Simulation-> simulate 3D (or 2D) and click the Advanced button.
You can increase the number of PML layers to further increase the overall absorption of the boundary (the maximum number is 100). Be aware that the more layers, the longer the simulation and the more memory needed. More layers will also increase reflections from glazing angles in some cases.
For really small structures and/or metallic structures, you can also increase the tensor parameter to around 80-85. Use at your own risk since it can also increase reflections in some cases!
Details are available in the documentation (technical background manual).
- You must be logged in to reply to this topic.