Backscattering simulation

[smgomez1995]

Dear Giovanni

Is there a way to do a simulation for the backscattering? Also, do you have a suggestion to run simulation using big spheres (20 um - 60 um diameter), 100 spheres? I am running them but it takes a lot of time.

Bests Sebastian

[gevero]

Hi Sebastian

Your simulation won't be trivial to run, if possible at all:

• a 60um sphere in the visible range requires at lest 100 multipole orders, according to the Wiscombe criterion
• Since you have interacting spheres, you need to calculate Vector translation coefficients (VTC) up to 100th order. I don't think that in the current implementation this can work. I think that, at most, I calculated them up to the 40th order or something like that. More than that, you are going to see numerical overflow due to the spherical hankel functions.
• If your spheres are kind of far away, one could truncate the VTC to the dipolar or quadrupolar order, but this is not implemented. it shuould not be hard to do it, but I am kind of busy with other projects now. I could guide you to do so if you wish.
• regarding the back scattering calculation, the subroutine "Efar" computes the radiation pattern, so you could get what you need from there, I guess.

[smgomez1995]

Hello Giovanni

Very thank you much for your response. In this case, I am more interested in what happens in the Near Infrared Range - Far Infrared Range (780 nm - 25 um). I understand that the scenario is unusual, so I am ok if the simulation just gives me tendencies (when I change the sphere distribution) than the actual numbers. My respond on your answers:

• a 60um sphere in the visible range requires at lest 100 multipole orders, according to the Wiscombe criterion. ok
• Since you have interacting spheres, you need to calculate Vector translation coefficients (VTC) up to 100th order. I don't think that in the current implementation this can work. I think that, at most, I calculated them up to the 40th order or something like that. More than that, you are going to see numerical overflow due to the spherical hankel functions. I understand that this is a limitation that I can not work around, right?
• If your spheres are kind of far away, one could trucate the VTC to the dipolar or quadrupolar order, but this is not implemented. it shuould not be hard to do it, but I am kind of busy with other projects now. I could guide you to do so if you wish. what will be the definition of far away? in this case, the average distance between spheres is 300 um. If thats the case, I will be happy to do the implementation with your guidance.
• regarding the back scattering calculation, the subroutine "Efar" computes the radiation pattern, so you could get what you need from there, I guess. I will explore it

Bests and thank you for your cooperation. Sebastian G.

[gevero]

Hi Sebastian

So, to resume, this is the situation:

• You have a cluster of about 100 sphere, 20-60um in diameter, with an average distance of 300um, so the distance is 5 to 10 times the sphere size. Furthermore, in a worst case scenario, your spheres are much larger than the incident wavelength. These are all the informations we need.
• The bad news is that, in the current form, the code is unable to handle your problem
• The good news is that, with the modification I suggested, the code should be able to easily handle your problem. The idea is that, since your spheres are relatively far away, only the first few orders of the scattered field of one sphere (dipole, quadrupole etc...) must be translated and transformed as an incident field on all the others spheres. This means that instead of calculating the VTC up to the 100th order, maybe we need at most the first few orders (something like 4 or 5 orders might be ok)
• Unfortunately I did not touch che fortran code in a long time, so the modification, either made by me or by you with my guidance, might take some time. What is your time frame?

Best

Giovanni

[smgomez1995]

Dear Giovanni I would like to have some good calculations in early March.

[smgomez1995]

Dear Giovanni,

I have reviewed your Fortran code and translated it to Python in order to better understand what it is doing. I also considered testing every function to double check the logic. I am currently 50% of the way through the translation (I am only translating what I need, not the entire code), and I think I could make some changes to the Fortran code with some effort.

Additionally, after some measurements of the phenomena I am studying, I have realized that:

The far infrared region may not be important to me anymore. I will confirm this after receiving the results of the measurements, so I will only be using the visible-near infrared range (300 nm up to 2500 nm). The spheres are generally very far apart, but sometimes they can be close to each other, with an average separation of around 300 um. I am extending my time frame by an additional 3 months. I believe I am capable of making the necessary modifications if you can provide some literature on the subject and some comments on it.

Any help you can provide would be greatly appreciated.

Best regards, Sebastian G.

[gevero]

Hi Sebastian

Sorry for the late reply, but it is a busy period. I gave a look at the code and I hope that the VTC truncation could be implemented rather easily. In the following day I hope I will be able to provide you with some directions.

[smgomez1995]

thank you Giovanni, I will wait for you instructions. Bests