2D to 3D simulations in FDTD when starting with effective index values

dcarlsondcarlson Member Posts: 1

I've been going through the 2D Topology Optimization example for a 4ch demultiplexer and am generally wondering how best to verify the output of this type of design with a 3D simulation. The material index used in the example is based on the effective index of an SOI slab mode, but in the 3D simulation the structure should be extruded using the actual material index instead of the slab mode index. The optimized region is an "Import" type object and I don't see how to scale the n,k values appropriately. Is there a better way to do this than exporting the structure to gds and re-importing? Thanks!


  • trobertstroberts Posts: 107Ansys Employee

    Hello @dcarlson ,

    For topology optimization the assumption that the slab modes are orthogonal, is not well preserved due to the nature of the structure. For the types of curves that are produced in shape optimization you find that the 2.5D approximations hold much better. It should be considered important to run the 3D optimization continuing from the 2D results, using the method outlined at the end of that article "Optimization Continuation"  Topology Optimization. Then set the max eps equal to the material index. Please let me know if you have any further questions.

    Best Regards,

Sign In or Register to comment.