Photonics – Korean

Photonics – Korean

FDE 시뮬레이션 관련 문의드립니다.

    • Choyunho
      Subscriber

      안녕하세요.

      FDE solver를 통해 mode를 구해보았습니다.

      첨부한 이미지를 보시면 loss term이 (-)가 있고 (+)가 있는데 (-)의 의미가 무엇인지 알 수 있을까요?

      그리고 varFDTDModeSource에서 구한 effective index와의 차이는 무엇인가요?

      감사합니다.

    • Guilin Sun
      Ansys Employee
      HI, Since Shin-Sung is taking more managemental responsibilities, I will reply you in English with the help of google translation.
      Your first question is about the sign of the imaginary part in neff. In general in Lumerical convention, "+" sign means it has loss whereas "-" sign means it has gain. However, for very small values such as 1e-5 or smaller (depending on the mesh, simulation region and boundary conditions", we consider it as numerical error. In particular, if the material has no gain, there is no physical base to have "-" sign other than numerical error.
      If necessary this value can be reduced with careful settings, such as using larger simulation region, finer mesh etc. Since the material refractive index has limited accuracy, in general no need to pursue much higher accuracy.
      Your second question seems like: what is the difference from the effective index obtained from varFDTDModeSource?
      As you can see, the current waveguide has a 2D cross section in FDE. However, in varFDTD, such 2D cross section has been simplified to be two 1D slabs: the 1st 1D-slab along z axis is through the green line at the waveguide core. Its neff is used as the refractive index of the "core" in 2D varFDTD, and its mode fields are used to calculate the effective index anywhere else. This is why the location of the green line is important. Detailed physics can be found here: https://support.lumerical.com/hc/en-us/articles/360034917213-MODE-2-5D-varFDTD-solver-introduction-
      The other 1D section is where you place the mode source in varFDTD, either along x or along y. Since varFDTD uses the 1D z slab information to get the equivalent refractive index for the simulation, the neff from the 1D mode source is close to the original 2D effective index.
      Those two 1D mode profiles can be combined to represent the original 2D mode profile. In other words, the original 2D mode profile is separated in the two 1D slab modes.
      In short, the neff and the mode profile from the 1D mode source in varFDTD represents the 2D waveguide in FDE within certain accuracy.

    • Choyunho
      Subscriber
      Thank you for your comment.
      But, my first question is meaning of (-)sign of loss[dB/cm] not imaginary part in neff.
      And I'll ask to make sure I understand the second answer.
      What method should be used to find the effective index of the high and low index parts of the grating structure?

    • Shin-Sung Kim
      Ansys Employee
      Loss 칸이 음수라는 것은 gain 이 발생한다는 것을 의미합니다. Loss 와 imaginary index 의 부호는 같습니다. 이 둘의 관계는 다음 페이지를 참고하시기 바랍니다.

      https://support.lumerical.com/hc/en-us/articles/360034917493-Working-with-lossy-modes-and-dB-m-to-kappa-conversion

      Gain material 을 사용하지 않았는데, 이런 unphysical 한 현상이 발생하는 것은, weakly-confined 된 mode 의 경우 pml boundary 와 구조물 간의 간격이 충분히 크지 않았기 때문일 가능성이 높습니다. 시뮬레이션의 x-span 을 더 늘려보시기 바랍니다. 그리고, Linear scale 로 visualize를 했을 때에는 mode 가 잘 confine 된 것으로 보이지만, log scale 로 보면 mode 가 얼마나 멀리까지 spread 하는 지 보다 잘 볼 수가 있습니다. Log plot 을 해서 pml boundary 주변의 field 의 intensity 가 중심부의 값에 대해 약 10^-10 정도까지 떨어지도록 x-span 을 잡아주면 적절히 시뮬레이션 영역을 잡았다고 볼 수 있습니다.

       

      graing waveguide 에서 high/low index region 의 각각의 위치에서의 effective index 값을 구하고자 한다면 FDE simulation region 의 위치를 바꾸어서 각 위치에서의 mode 를 계산해주면 됩니다.
Viewing 3 reply threads
  • You must be logged in to reply to this topic.