Questions and Answers

0 Like 0 Dislike


Value Error Message?


I am wondering if nanoDDSCAT is appropriate for modeling waveguides with nanoparticles on the surface. Every time I attempt to model a rectangular waveguide with nanoparticle shapes on the surface, incident light along the x axis, and run the DDA Convert, I receive the message: "ValueError: invalid literal for int() with base 10" Is this because my parameters were too large? I tried with a 1-micron long, 365 nm width waveguide, then decreased it below that, but I still keep getting this message, and I am not sure if this is because of my parameters, or because the program is not fit for this type of application.

Report abuse

1 Responses

  1. 0 Like 0 Dislike

    Prashant K Jain

    Thanks for using nanoDDSCAT. Let me attempt to provide some suggestions.

    First, without knowing the details of your simulation structure and how you are creating it, it is difficult to exactly pinpoint the source of error. But, it appears from the error message that DDAConvert is unable to open the file. Most likely this is because the file does not exist, or it has a different name than what DDAconvert expects, or the file is not loaded into the working memory.

    I would suggest using default options, where you specify that the obj file generated by Blender is directly employed as the input for DDAConvert. You should verify the obj file is indeed being created. If it is not being created, it could mean the problem is with the structure generation in Blender. This can happen if Blender encounters any errors in discretizing the object. If your object is too large, I would try testing with a smaller, simpler object, e.g. a 20 nm sphere.

    As for your specific target (1-micron long, 365 nm width waveguide), if you are using 1 dipole/nm density, then your dipole mesh has a 1000 * 365 * 365 size, which is rather large. The remedy is to use a sparser dipole density, say 1 dipole/10 nm. The waveguide would still have the same dimensions that you want to simulate, but the discretization of the waveguide structure would be coarser. There will be some hit to the accuracy but given how large your structure is, I doubt you need to have a 1 dipole/nm density.

    Best wishes,



    Reply Report abuse

    Please login to answer the question.