How large size parameters can this code calculate？

[sunknightlee]

Hello, I am using this code to calculate scattering coefficient and extinction coefficient. But I find that the execution of the code stops when I try to calculate the T-matrix of a scatterer with a size parameter which is about 80 while the axis_ratio = 1, and the size parameter is reduced to about 18 while the axis_ratio = 3. Is this a bug? Or it's a existing limit of the fortran numerical code? According to table 1 in the paper named "Capabilities and limitations of a current FORTRAN implementation of the T-matrix method for randomly oriented, rotationally symmetric scatterers" by M. I. Mishchenko, the maximal convergent size parameters seem to be much larger than the limit I find. By the way, I have tried to increase value of the attribute ndgs in the Scatterer class, but still I can't calculate the T-matrix when size parameter is large. So, I want to know how large size parameters can this code calculate? And what should I do to calculate T-matrix for larger size parameter? I'm looking forward to your reply. Thanks for your kindness! Best, Lee

[jleinonen]

Hi Lee,

Convergence issues seem fairly typical at these size parameters in my experience. What refractive index did you use? That will also affect convergence.

[sunknightlee]

Yeah, refractive index also has effects. If I use refractive index 1.311 like I mentioned in the previous table，and use axis_ratio = 3, I can calculate until the size parameter is around 20. For larger size parameter, the code will terminate. This value is significantly smaller than that in Table 1.