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 The information provided may be out of date.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.

Time dependant Structural Mechanics & Magnetic Field

Please login with a confirmed email address before reporting spam

Hy

Can anyone please let me know that is it possible to simulate time dependant magnetic field when force or some external conditions are there in which geometrical position of the shape is changing.
Comsol builtin model of freely falling magnet is one case of Time dependant magnetic field but is the above case mentioned by me applicable? And what if I combine moving mesh together and simulate all three physics together? Which studies I should carry out.

Its important for me, plz someone help.
Thanks in advance.
Regards
MF

14 Replies Last Post 3 avr. 2014, 10:14 UTC−4

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 3 janv. 2014, 08:38 UTC−5
Hi,

As far as I can say, if your moving domain has to interract (in both directions) with the magnetic flux at different displacements, then you should capture that by introducing moving mesh (ale). If the interaction goes in single direction or if it has a pattern (the variables are dependent on each other), then you should define it as a position/time dependent variable and try to avoid complexity.

A simple way to construct such model is to use Magnetic Fields (mf) module in combination with ALE and Global ODE/PDEs (ge). You define the forces in (mf) and you define how the forces are acting on your moving part(s) in (ge) by defining the equation of motion and let the (ale) take care how the geometry displaces in your model.

Hope it helps,
Onur
Hi, As far as I can say, if your moving domain has to interract (in both directions) with the magnetic flux at different displacements, then you should capture that by introducing moving mesh (ale). If the interaction goes in single direction or if it has a pattern (the variables are dependent on each other), then you should define it as a position/time dependent variable and try to avoid complexity. A simple way to construct such model is to use Magnetic Fields (mf) module in combination with ALE and Global ODE/PDEs (ge). You define the forces in (mf) and you define how the forces are acting on your moving part(s) in (ge) by defining the equation of motion and let the (ale) take care how the geometry displaces in your model. Hope it helps, Onur

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 3 janv. 2014, 18:17 UTC−5
Hy Onur Ilkorur thanks a lot lot for so kind reply, its ur great favor as I have been struggling a lot from 3 months.

I am confused about that, Plz have an overview of the attached file, this can make my case clear:
Plz plz guide me that which "Physics+Study" should be carried out for the voltage induction from a rectangular beam of a magnet (parallel to x axis ) to a simple coil (parallel to y axis)? The beam is fixed at one end.

What I am trying is moving mesh, solid mechanics for fixing one side of beam and applying force for displacement and magnetic field physics (ampere's law & multi-turn linear coil with reference edge).

Which domains I should use in moving mesh (if it is required)? In output I want a SINUSOID.

And last querry please Comsol inquires about Young's Modulus and Poisons Ratio for air? What values should be used?

Waiting for your kind reply.
Best Regards
MF
Hy Onur Ilkorur thanks a lot lot for so kind reply, its ur great favor as I have been struggling a lot from 3 months. I am confused about that, Plz have an overview of the attached file, this can make my case clear: Plz plz guide me that which "Physics+Study" should be carried out for the voltage induction from a rectangular beam of a magnet (parallel to x axis ) to a simple coil (parallel to y axis)? The beam is fixed at one end. What I am trying is moving mesh, solid mechanics for fixing one side of beam and applying force for displacement and magnetic field physics (ampere's law & multi-turn linear coil with reference edge). Which domains I should use in moving mesh (if it is required)? In output I want a SINUSOID. And last querry please Comsol inquires about Young's Modulus and Poisons Ratio for air? What values should be used? Waiting for your kind reply. Best Regards MF


Sergei Yushanov Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 6 janv. 2014, 12:27 UTC−5
Mahmood,

Looking at the screenshot of your mode, I assume that entire coil is fixed, beam is cantilever and can move due to Lorentz force. Based on these assumptions, several notes on your model:
1. All domains should be selected in Magnetic Fields (mf2) interface;
2. Only beam should be selected in Solid Mechanics (solid) interface;
3. Only air domain should be selected in Moving Mesh (ale2) interface. Young’s modulus and Poisson’s ratio for air are not needed. Comsol is asking these properties because you included air domain in Solid mechanics interface.
4. Beam movement is governed by Solid Mechanics interface, so it is not necessary to include beam in Moving Mesh interface. However, use Prescribed Mesh Displacement boundary condition in Moving Mesh interface to set displacement of the beam boundaries to (u,v,w).
5. Apply Maxwell’s stress tensor forces on all boundaries of the beam in Solid Mechanics interface. This is the coupling between Magnetic Fields and Solids Mechanics.
6. You coil looks very strange: where is closed loop for current flow?
7. You can get away with only one Time Dependent study step.

Best regards,
Sergei
Mahmood, Looking at the screenshot of your mode, I assume that entire coil is fixed, beam is cantilever and can move due to Lorentz force. Based on these assumptions, several notes on your model: 1. All domains should be selected in Magnetic Fields (mf2) interface; 2. Only beam should be selected in Solid Mechanics (solid) interface; 3. Only air domain should be selected in Moving Mesh (ale2) interface. Young’s modulus and Poisson’s ratio for air are not needed. Comsol is asking these properties because you included air domain in Solid mechanics interface. 4. Beam movement is governed by Solid Mechanics interface, so it is not necessary to include beam in Moving Mesh interface. However, use Prescribed Mesh Displacement boundary condition in Moving Mesh interface to set displacement of the beam boundaries to (u,v,w). 5. Apply Maxwell’s stress tensor forces on all boundaries of the beam in Solid Mechanics interface. This is the coupling between Magnetic Fields and Solids Mechanics. 6. You coil looks very strange: where is closed loop for current flow? 7. You can get away with only one Time Dependent study step. Best regards, Sergei

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 7 janv. 2014, 07:43 UTC−5
Hy Honorable Sirgei Yushanov, hope you are doing good.

First of all off course millions of thanks for so kind and detailed reply. I implemented your instructions as suggested by you.
I have last 3 queries, which I have elaborated in attached word file.
Kindly and please spare your precious time and have a look at that word file and help me in solving that issue as well, it will help a lot lot in future and it will definitely be yours memorable favor. Kindly e-mail me at the mentioned address or write here at blog.
Thanks in advance and waiting for your kind help please.

Regards
M. Farman (m.farman134@yahoo.com)
Hy Honorable Sirgei Yushanov, hope you are doing good. First of all off course millions of thanks for so kind and detailed reply. I implemented your instructions as suggested by you. I have last 3 queries, which I have elaborated in attached word file. Kindly and please spare your precious time and have a look at that word file and help me in solving that issue as well, it will help a lot lot in future and it will definitely be yours memorable favor. Kindly e-mail me at the mentioned address or write here at blog. Thanks in advance and waiting for your kind help please. Regards M. Farman (m.farman134@yahoo.com)


Sergei Yushanov Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 7 janv. 2014, 12:23 UTC−5
Mahmood,

1. Material parameters for air are OK. Generally, small conductivity of air is needed to resolve solver convergence issues. Start with some relatively small air electrical conductivity. If solver converges, you can try to further reduce air conductivity.
2. Default magnetic insulation boundary conditions at the exterior boundaries are OK as long as exterior boundaries are not too close to the structure of interest.
3. Don’t change default Prescribed Mesh Displacement node. Add Free Deformation node for air domain and then add another Prescribed Mesh Displacement node for the boundaries of the beam where you set Prescribed x-, y-, and z-displacements as (u,v,w).
4. Variables for boundary load due to Maxwell’s stress tensor are: (mf.unTx, mf.unTy, mf.unTz,).
5. You might need to change default solver setting for Jacobian update from “minimal” to “On every iteration” if you encounter convergence difficulties.
I am attaching simple test model as an illustration of problem setup. Please note that geometry and material parameters are arbitrary and has nothing to do with real problem.

Best regards,
Sergei
Mahmood, 1. Material parameters for air are OK. Generally, small conductivity of air is needed to resolve solver convergence issues. Start with some relatively small air electrical conductivity. If solver converges, you can try to further reduce air conductivity. 2. Default magnetic insulation boundary conditions at the exterior boundaries are OK as long as exterior boundaries are not too close to the structure of interest. 3. Don’t change default Prescribed Mesh Displacement node. Add Free Deformation node for air domain and then add another Prescribed Mesh Displacement node for the boundaries of the beam where you set Prescribed x-, y-, and z-displacements as (u,v,w). 4. Variables for boundary load due to Maxwell’s stress tensor are: (mf.unTx, mf.unTy, mf.unTz,). 5. You might need to change default solver setting for Jacobian update from “minimal” to “On every iteration” if you encounter convergence difficulties. I am attaching simple test model as an illustration of problem setup. Please note that geometry and material parameters are arbitrary and has nothing to do with real problem. Best regards, Sergei


Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 9 janv. 2014, 00:35 UTC−5
Hy Sir Sergei, first of all I am extremely thankful for such a big and kind favor of yours and it gave a lot of happiness, as it was almost 90 percent of the task which I was striving for and this memorable help really wouldn't have been possible without your help. Immense thanks again sir.

I was implementing it myself from yesterday and so went through a query about the Maxwell stress because the task which I have to do is to Vibrate that beam at some particular frequency and then induce flux in the coil whereas in the instructions it was done by Maxwell stress body load and hopefully if I move the coils under the beam, voltage will still be induced?

And can I increase that voltage by Halbach array formation of magnets ad finally how to convert that voltage into power (watts) and it will be my task.

If possible kindly attach some model by which I can go ahead and this is the last favor for which I am taking your precious time.

Immense thanks.
Best wishes.
MF
Hy Sir Sergei, first of all I am extremely thankful for such a big and kind favor of yours and it gave a lot of happiness, as it was almost 90 percent of the task which I was striving for and this memorable help really wouldn't have been possible without your help. Immense thanks again sir. I was implementing it myself from yesterday and so went through a query about the Maxwell stress because the task which I have to do is to Vibrate that beam at some particular frequency and then induce flux in the coil whereas in the instructions it was done by Maxwell stress body load and hopefully if I move the coils under the beam, voltage will still be induced? And can I increase that voltage by Halbach array formation of magnets ad finally how to convert that voltage into power (watts) and it will be my task. If possible kindly attach some model by which I can go ahead and this is the last favor for which I am taking your precious time. Immense thanks. Best wishes. MF

Sergei Yushanov Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 9 janv. 2014, 10:20 UTC−5
Mahmood,

Seems that my understanding of the problem you are trying to solve was incorrect. Based on your latest description, you need to set up the problem the other way around: prescribe sinusoidal motion to the magnet and calculate induced voltage in the coil. This is somewhat simpler problem since you don’t need Solid Mechanics interface and no Maxwell’ stress tensor is required: just prescribe magnet motion in the Moving Mesh interface. Then, induced voltage can be calculated by integrating electric field over cross-section of the coil along the coil circumferential loop.

Best regards,
Sergei
Mahmood, Seems that my understanding of the problem you are trying to solve was incorrect. Based on your latest description, you need to set up the problem the other way around: prescribe sinusoidal motion to the magnet and calculate induced voltage in the coil. This is somewhat simpler problem since you don’t need Solid Mechanics interface and no Maxwell’ stress tensor is required: just prescribe magnet motion in the Moving Mesh interface. Then, induced voltage can be calculated by integrating electric field over cross-section of the coil along the coil circumferential loop. Best regards, Sergei

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 9 janv. 2014, 13:09 UTC−5
Thanks a lot sir for your guidance. I really am hesitating to take your time again and again. Just give me that last favour of yours and please please make changes in the above attached file by you at any PARTICULAR FREQUENCY of vibration because it will again take a lot of time of mine as it took already about 3 months of mine and it wouldn't have been possible without you sir.
Because how should I assign magnet motion in terms of frequency and how will I fix one end of it please (probably by not selecting one of the boundry)?
And sorry I am totally unaware of integrating as told by you.

Please please make changes in the attached file and my work will be done, it will be your great magnanimity please and last favour please because I already took too much of your time.
Waiting for you kind reply please and thanks again for your so so kind help.
Best Regards sir,
MF
Thanks a lot sir for your guidance. I really am hesitating to take your time again and again. Just give me that last favour of yours and please please make changes in the above attached file by you at any PARTICULAR FREQUENCY of vibration because it will again take a lot of time of mine as it took already about 3 months of mine and it wouldn't have been possible without you sir. Because how should I assign magnet motion in terms of frequency and how will I fix one end of it please (probably by not selecting one of the boundry)? And sorry I am totally unaware of integrating as told by you. Please please make changes in the attached file and my work will be done, it will be your great magnanimity please and last favour please because I already took too much of your time. Waiting for you kind reply please and thanks again for your so so kind help. Best Regards sir, MF

Sergei Yushanov Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 10 janv. 2014, 09:52 UTC−5
Mahmood,

Although I am not sure what you are trying to achieve, the attached is modified file with beam forced vibration.

Best regards,
Sergei
Mahmood, Although I am not sure what you are trying to achieve, the attached is modified file with beam forced vibration. Best regards, Sergei


Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 10 janv. 2014, 12:12 UTC−5
Immense thanks sir for your kind favour and precious time. If you are interested, kindly do share your email and I can tell you the details about my idea on which I have done literature review and want to work on, might it be helpful to you as well as many of the persons have appreciated my idea. And after my work in 2 to 3 months, it will be an honor for me to share it with you as it would have been totally impossible without you.

And off course thanks for the beam forced vibration and the time dependent study, it is exactly in accordance to what I was striving for. And in the output of induced voltage there is a straight line (file attached) at zero but in previous case of Maxwell stress it was a sinusoid. I am confused in that, please guide about that.

Best Regards,
M. Farman (m.farman134@yahoo.com)
Immense thanks sir for your kind favour and precious time. If you are interested, kindly do share your email and I can tell you the details about my idea on which I have done literature review and want to work on, might it be helpful to you as well as many of the persons have appreciated my idea. And after my work in 2 to 3 months, it will be an honor for me to share it with you as it would have been totally impossible without you. And off course thanks for the beam forced vibration and the time dependent study, it is exactly in accordance to what I was striving for. And in the output of induced voltage there is a straight line (file attached) at zero but in previous case of Maxwell stress it was a sinusoid. I am confused in that, please guide about that. Best Regards, M. Farman (m.farman134@yahoo.com)


Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 11 janv. 2014, 14:16 UTC−5
Hy Sir Sergei Yushanov...

Sir please guide me about that last issue, its a very humble request and I will not disturb you after that. It really is very very important for me.
Just guide me about the Sinusoid output of VOLTAGE which a straight line in the last file send by you. Is voltage induced or I am getting it wrong? Whereas Vibrations are perfect and their sinusoid is there.

Please give me that last favour about Sinuoid of VOLTAGE output (which you shared by Maxwell's stress was perfect but it was by Stress). And my work will be done.
And again apology for disturbing you so much but please do respond.

Waiting for your kind response please.
Hy Sir Sergei Yushanov... Sir please guide me about that last issue, its a very humble request and I will not disturb you after that. It really is very very important for me. Just guide me about the Sinusoid output of VOLTAGE which a straight line in the last file send by you. Is voltage induced or I am getting it wrong? Whereas Vibrations are perfect and their sinusoid is there. Please give me that last favour about Sinuoid of VOLTAGE output (which you shared by Maxwell's stress was perfect but it was by Stress). And my work will be done. And again apology for disturbing you so much but please do respond. Waiting for your kind response please.


Sergei Yushanov Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 13 janv. 2014, 10:18 UTC−5
Mahmood,

Attached is model with induced sinusoidal voltage of the coil.

Best regards,
Sergei
Mahmood, Attached is model with induced sinusoidal voltage of the coil. Best regards, Sergei


Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 13 janv. 2014, 23:34 UTC−5
Hy Sir Sergei Yushanov...

Hope you doing good. Sir you are simply a great and a magnificent person I have ever been with. Bundle of thanks again for this magnanimity of yours please.
And please if u wanna share your personal email, it will be an honor for me to be get your final approval in about two months for a design task on which I am working for or kindly do email me.
Forever thankful to you for the spreading of knowledge as well for this forum and we novice people in this field, you are one of the truly dedicated person.

Best Regards,
Your's unknown student :)
M. F (m.farman134@yahoo.com)
Hy Sir Sergei Yushanov... Hope you doing good. Sir you are simply a great and a magnificent person I have ever been with. Bundle of thanks again for this magnanimity of yours please. And please if u wanna share your personal email, it will be an honor for me to be get your final approval in about two months for a design task on which I am working for or kindly do email me. Forever thankful to you for the spreading of knowledge as well for this forum and we novice people in this field, you are one of the truly dedicated person. Best Regards, Your's unknown student :) M. F (m.farman134@yahoo.com)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 3 avr. 2014, 10:14 UTC−4
Dear mahmood and Sergei,
i'm a student from electrical enginerring, and i'm stuck in this simple problem: how can you apply lorent'z forces and maxwell's stress tensor to the solid model domain, in my case a molten metal!

Thanks in advance!
Gianmarco
Dear mahmood and Sergei, i'm a student from electrical enginerring, and i'm stuck in this simple problem: how can you apply lorent'z forces and maxwell's stress tensor to the solid model domain, in my case a molten metal! Thanks in advance! Gianmarco

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.