A word of caution on simulations...
Personally I’m not a big fan of doing simulations, I solely use them to verify an idea of how something is behaving. That is to say, I have a some qualitative and quantitate knowledge of what I’m attempting to simulate. From the mentioned software packages, I only have experience using COMSOL Multiphysics. This software uses classical equations to solve most problems which for most cases its sufficient. However, when performing simulations in the nano scale you should exercise caution in interpreting your results. If say you want to use some mesoscopic equations, you can set up your own equations and define parameters like boundary conditions which you can apply to your model. This saves you the hassle of typing your own code and using descent CAD system and FEM solver to perform your calculations.
Whenever you have new code, always test it out with things that are well understood before attempting to invest time in things which you really don’t know how they function.
Just wanted to share my thoughts. I’m not stating they are not useful... but to proceed with caution when interpreting you results.
Thanks for recommendations on the other software packages.
Cheers,
Sergio