Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.
Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
Scatterer on substrate, RF module
Posted 13.04.2013, 18:12 GMT-4 Version 4.2a, Version 4.3a, Version 4.3b 3 Replies
Please login with a confirmed email address before reporting spam
Dear All,
Probably some of you RF module users saw module “Scatterer on substrate” in model gallery.
I have question about that, and I have question about total field formalism in generally.
So, basic idea is to start simulation with two calculations. First for calculating background field when we don't have scatterer on substrate, and second for calculating scattered field when we have scatterer on substrate and where we have background field from first calculation as incident field.
My first question, is there possible to solve these problems with two separate simulations???
So I want to start first simulation just for calculation of background field, and after that somehow to impose results from first simulation in second simulation where we want to find scattered field. So I don’t want to have two calculations in one simulation, but to separate them.
My second question (if is not possible to solve problem from my first question), is there possible to solve background field but without quality (or good) meshing of the scatterer. As you can see from that new module, in first calculation we have to define scaterrer like an air, but still there is a many mesh element, because we meshed that part of space on proper way for next simulation where we want to find scattered field. So, is there possible in first calculation of background field somehow to avoid good meshing of that part of space, and just to have good meshing in second calculation, where we really want to calculate scattered properties.
Maybe you already realize that my questions are related with saving time and memory, so I want to find out is there any possibility to solve these problems with much less mesh elements in first calculation, and with proper number of mesh elements in second calculation.
Thanks in advanced for every suggestion and answer.
Cheers.
Probably some of you RF module users saw module “Scatterer on substrate” in model gallery.
I have question about that, and I have question about total field formalism in generally.
So, basic idea is to start simulation with two calculations. First for calculating background field when we don't have scatterer on substrate, and second for calculating scattered field when we have scatterer on substrate and where we have background field from first calculation as incident field.
My first question, is there possible to solve these problems with two separate simulations???
So I want to start first simulation just for calculation of background field, and after that somehow to impose results from first simulation in second simulation where we want to find scattered field. So I don’t want to have two calculations in one simulation, but to separate them.
My second question (if is not possible to solve problem from my first question), is there possible to solve background field but without quality (or good) meshing of the scatterer. As you can see from that new module, in first calculation we have to define scaterrer like an air, but still there is a many mesh element, because we meshed that part of space on proper way for next simulation where we want to find scattered field. So, is there possible in first calculation of background field somehow to avoid good meshing of that part of space, and just to have good meshing in second calculation, where we really want to calculate scattered properties.
Maybe you already realize that my questions are related with saving time and memory, so I want to find out is there any possibility to solve these problems with much less mesh elements in first calculation, and with proper number of mesh elements in second calculation.
Thanks in advanced for every suggestion and answer.
Cheers.
3 Replies Last Post 04.01.2014, 07:01 GMT-5
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
hi all,
i am also interesting to know the answer as well.
thanks
i am also interesting to know the answer as well.
thanks
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
I am by no means an expert in comsol, but I think I found an answer to your problem, which you may have figured out by now. However, I thought this might be useful for someone else.
If you click on Step 1: frequency domain under "study 1" then scroll down you will see "mesh selection". I have never tried this but it appears that you could have two different mesh selections for each step (step 1 and step 2)....? Well at least this is the case in 4.3b, but I think 4.3a has this option... I was able to add ANOTHER mesh node and select it instead of the original, first mesh, so I think it may work for you...
It seems to me that this would not just save time and computer resources, but it would also just make more sense and give you a more accurate answer. Solving more points in the gold nanoparticle when it's defined as air compared to the surrounding air just doesn't make sense to me. Again, I am not an expert, so feel free to correct me if I am wrong.
If you click on Step 1: frequency domain under "study 1" then scroll down you will see "mesh selection". I have never tried this but it appears that you could have two different mesh selections for each step (step 1 and step 2)....? Well at least this is the case in 4.3b, but I think 4.3a has this option... I was able to add ANOTHER mesh node and select it instead of the original, first mesh, so I think it may work for you...
It seems to me that this would not just save time and computer resources, but it would also just make more sense and give you a more accurate answer. Solving more points in the gold nanoparticle when it's defined as air compared to the surrounding air just doesn't make sense to me. Again, I am not an expert, so feel free to correct me if I am wrong.
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hello,
I am also going throw this simulation. I am testing the script and I found that for ~500nm wavelength (600 THz) it converges very quickly (less than 2 min on my computer). But when I try to simulate a 200nm wavelength it takes forever to converge (I am still waiting for it to converge....)
Any ideas how to solve this problem for smaller wavelengths?
I am also going throw this simulation. I am testing the script and I found that for ~500nm wavelength (600 THz) it converges very quickly (less than 2 min on my computer). But when I try to simulate a 200nm wavelength it takes forever to converge (I am still waiting for it to converge....)
Any ideas how to solve this problem for smaller wavelengths?