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.

Sphere-Sphere Forces

Please login with a confirmed email address before reporting spam

Hello. Firstly, greetings to everyone, I am new here.

So, I am trying to build a model based on this article:
www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&cad=rja&ved=0CCoQFjAB&url=http%3A%2F%2Fwww.victoria.ac.nz%2Fscps%2Fabout%2Fstaff%2Fpdf%2FElectrostatic-calibration-of-spheresphere-forces.pdf&ei=jF19UIPEL9ORhQfbsYC4AQ&usg=AFQjCNHvsG40c8nlNlCR3yx7YKBfey-tXw&sig2=HC5XPGRvo0qippVGnfngLg

I am also new to electrostatic modeling, but I have some basic Comsol knowledge.

My model contains two copper spheres with potentials Va and Vb and a large sphere representing surrounding medium (vacuum). I have selected Force Calculation domain condition and I tried Electric Potential and Terminal boundary conditions on the spheres.

In the article, it is described: '' When Va=Vb the force is repulsive at all distances, otherwise the force becomes attractive when the approach is close enough. Attraction occurs between spheres both at positive potential because of mutual polarization: for spheres with equal radii, the sphere at lower potential obtains a negatively charged region neighbouring the other sphere.''

I tried many different settings and I can't seem to get the described result, especially the reversal of the force direction. Am I missing something fundamental or just any detail? Is this a phenomenon that can be modeled using Comsol?

Regards,
Tomi

7 Replies Last Post Nov 5, 2012, 5:23 a.m. EST

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Oct 16, 2012, 12:21 p.m. EDT
Hello, Tomo

I suggest you to define an external (spherical) layer and there set Infinite Elements, and in the most external boundary set condition to Ground. You can set all the domains as air, and one Force Calculation domain condition per each little sphere. This way you'll be able to compute the forces.

Bye,

Jesus.
Hello, Tomo I suggest you to define an external (spherical) layer and there set Infinite Elements, and in the most external boundary set condition to Ground. You can set all the domains as air, and one Force Calculation domain condition per each little sphere. This way you'll be able to compute the forces. Bye, Jesus.

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Oct 17, 2012, 8:15 a.m. EDT
Thank you, Jesus. Now it seems to be working correctly.

With Infinite Elements layer the surrounding medium domain can now be smaller, correct? What is the usual recommendation for the ratio? My surrounding medium sphere is now ten times the radius of the small sphere.

Bye,
Tomo
Thank you, Jesus. Now it seems to be working correctly. With Infinite Elements layer the surrounding medium domain can now be smaller, correct? What is the usual recommendation for the ratio? My surrounding medium sphere is now ten times the radius of the small sphere. Bye, Tomo

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Oct 17, 2012, 10:31 a.m. EDT
Hello, Tomo,

You are right. I don't know if there is any recommendation for the ratio between spheres radius. I think ten is good choice.

Jesus.
Hello, Tomo, You are right. I don't know if there is any recommendation for the ratio between spheres radius. I think ten is good choice. Jesus.

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Oct 17, 2012, 2:14 p.m. EDT
Hi
For decaying fields I use about 1/ten ratio, for waves I used a few wavelength minimum (anyhow these stretch as the inf elements goes out). It is discussed in the doc somewhere, just cant remember eactly where, try a search

--
Good luck
Ivar
Hi For decaying fields I use about 1/ten ratio, for waves I used a few wavelength minimum (anyhow these stretch as the inf elements goes out). It is discussed in the doc somewhere, just cant remember eactly where, try a search -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Oct 31, 2012, 10:55 a.m. EDT
Hello again,

I have successfully recreated the desired results with only one difference. They are off by a factor of 2, i.e. twice as big as in the article. In the attached picture are my results(divided by 2) photoshoped over the graph from the article. They align as they should.
Now I cannot find the reason for the factor of two. I tried rebuilding the model, different parameters, mesh sizes; the results are always the same. Then I tried to calculate the force manually via integration of the Maxwell Stress Tensor: the same.
We contacted the author of the article, and he suggested we calculate the forces for a large separation between spheres. We did and we compared them to a calculation for a force between two charges, which should be a good enough approximation for a large separation. Yes, they are also off and the results divided by 2 align perfectly.

Any ideas about this?

--
Regards,
Tomo
Hello again, I have successfully recreated the desired results with only one difference. They are off by a factor of 2, i.e. twice as big as in the article. In the attached picture are my results(divided by 2) photoshoped over the graph from the article. They align as they should. Now I cannot find the reason for the factor of two. I tried rebuilding the model, different parameters, mesh sizes; the results are always the same. Then I tried to calculate the force manually via integration of the Maxwell Stress Tensor: the same. We contacted the author of the article, and he suggested we calculate the forces for a large separation between spheres. We did and we compared them to a calculation for a force between two charges, which should be a good enough approximation for a large separation. Yes, they are also off and the results divided by 2 align perfectly. Any ideas about this? -- Regards, Tomo


Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Oct 31, 2012, 3:47 p.m. EDT
Hi

which version are you using?
if I remember right there was a factor 2 missing in some ACDC internal variables in the early 4.0, could it be that ?

--
Good luck
Ivar
Hi which version are you using? if I remember right there was a factor 2 missing in some ACDC internal variables in the early 4.0, could it be that ? -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Nov 5, 2012, 5:23 a.m. EST
I am using 4.3.
It seems that I have managed to eliminate the problem. I really don't know what the problem was, but I rebuilt the model once again and now the results are correct. I can not find the difference in the models, but one of the two gives the correct results and the other one does not. I must have made some mistake somewhere, because after another rebuild the results are still correct.
Anyway, thank you for your help. I will try to triple-check in the future before posting here.

Regards,
Tomo
I am using 4.3. It seems that I have managed to eliminate the problem. I really don't know what the problem was, but I rebuilt the model once again and now the results are correct. I can not find the difference in the models, but one of the two gives the correct results and the other one does not. I must have made some mistake somewhere, because after another rebuild the results are still correct. Anyway, thank you for your help. I will try to triple-check in the future before posting here. Regards, Tomo

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.