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 The information provided may be out of date.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.
3D Induction Motor Modelling
Posted Dec 12, 2011, 5:36 p.m. EST Heat Transfer & Phase Change Version 4.2 6 Replies
Please login with a confirmed email address before reporting spam
Hi guys
I am trying to model a 3D three phase induction motor. I have attached the model. I went through previous discussions in this forum and have tried my best to build the model. The geometries are built in SolidWorks and the imported in comsol as different entities, which are then combined together to make a rotating part (rotor) and a stationary part (stator). The final geometry is assembled using 'form an assembly' while finalizing. There is a continuity pair defined in the boundary between the stator and rotor.
Some domains(stator slots and air) is using magnetic permeability defined in the material, where as some domains (stator and rotor core) are using the HB Curve, user defined. Two separate Ampere's Law have been used for these.
I have created a mesh in a way that I first created edges in the boundary between airpgar, rotor bars and the boundary of the stator near the airpap. then free triangular mesh is created in individual domains, which are then swept across the other side of the geometry. There are no error shown in the mesh.
Now my questions are listed as follows:
1) I have used external current density as a source. The three phase current density are marked with Ja, Jb and Jc, where '+ve' and '-ve' in the parenthesis means the inward and outward current respectively (similar to the cross and dot convention). But I read somewhere in this forum that simpling defining current density in a domain, without showing the flow of current is not valid in 3D. Is that true? If so, how should I define the flow? Can you please check the way I have defined the current density. Is it right? Do I need to connect the conductor at the end side?
2) I am doing a time-dependent study. To study the rotation in a 2D dimensional model was easy. But how to rotate the rotor in 3D? Shall I use the 'Moving Mesh'?. How to use that?
3)Until now, I have been doing it without considering the rotation. But in that case, the simulation begins but doesnot continue. Even simulating for 0.5 sec, it does not complete. It just begins, and remains as it is. The convergenge graphs even does not turn up!
I would be very very thankful for the help. Or atleast for an idea, how to proceed.
Best Regards
I am trying to model a 3D three phase induction motor. I have attached the model. I went through previous discussions in this forum and have tried my best to build the model. The geometries are built in SolidWorks and the imported in comsol as different entities, which are then combined together to make a rotating part (rotor) and a stationary part (stator). The final geometry is assembled using 'form an assembly' while finalizing. There is a continuity pair defined in the boundary between the stator and rotor.
Some domains(stator slots and air) is using magnetic permeability defined in the material, where as some domains (stator and rotor core) are using the HB Curve, user defined. Two separate Ampere's Law have been used for these.
I have created a mesh in a way that I first created edges in the boundary between airpgar, rotor bars and the boundary of the stator near the airpap. then free triangular mesh is created in individual domains, which are then swept across the other side of the geometry. There are no error shown in the mesh.
Now my questions are listed as follows:
1) I have used external current density as a source. The three phase current density are marked with Ja, Jb and Jc, where '+ve' and '-ve' in the parenthesis means the inward and outward current respectively (similar to the cross and dot convention). But I read somewhere in this forum that simpling defining current density in a domain, without showing the flow of current is not valid in 3D. Is that true? If so, how should I define the flow? Can you please check the way I have defined the current density. Is it right? Do I need to connect the conductor at the end side?
2) I am doing a time-dependent study. To study the rotation in a 2D dimensional model was easy. But how to rotate the rotor in 3D? Shall I use the 'Moving Mesh'?. How to use that?
3)Until now, I have been doing it without considering the rotation. But in that case, the simulation begins but doesnot continue. Even simulating for 0.5 sec, it does not complete. It just begins, and remains as it is. The convergenge graphs even does not turn up!
I would be very very thankful for the help. Or atleast for an idea, how to proceed.
Best Regards
Attachments:
6 Replies Last Post Dec 24, 2015, 3:38 a.m. EST
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Anyways, the error now shows that, the solver did not find consistent initial values, out of memory while LU factorization.
And there is also a warning that, MUMPS allocation factor increases to 1.728.
And there is also a warning that, MUMPS allocation factor increases to 1.728.
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
in 3D you need a large RAM & CPU, for your case with such a symmetry you should start in 2D, once that is OK, then you could go to 3D. Where indeed you have 2 issues, the coil current looping, and the fringe effect of the B field looping around on the two sides, this is less trivial. But a 2D model would give you quite some info already
--
Good luck
Ivar
in 3D you need a large RAM & CPU, for your case with such a symmetry you should start in 2D, once that is OK, then you could go to 3D. Where indeed you have 2 issues, the coil current looping, and the fringe effect of the B field looping around on the two sides, this is less trivial. But a 2D model would give you quite some info already
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
HI
thanks for the reply! :)
I am using a computer with Windows 7, 64 bit OS with 8 GB RAM and Intel(R) Core(TM) 2 Duo processor. I guess this would work on this CPU.
I did not understand by what you meant '' Where indeed you have 2 issues, the coil current looping, and the fringe effect of the B field looping around on the two sides, this is less trivial''
Anyways, I have already done the 2D study of the machine. Now I want to see the behavior of the eddy currents in the rotor in 3D. Please help!
thanks for the reply! :)
I am using a computer with Windows 7, 64 bit OS with 8 GB RAM and Intel(R) Core(TM) 2 Duo processor. I guess this would work on this CPU.
I did not understand by what you meant '' Where indeed you have 2 issues, the coil current looping, and the fringe effect of the B field looping around on the two sides, this is less trivial''
Anyways, I have already done the 2D study of the machine. Now I want to see the behavior of the eddy currents in the rotor in 3D. Please help!
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
I hope it will go on your workstation, but Eddy currents is far more trickier too set up and to solve.
One of the issues in 3D is that the magnetic field must loop around, that is that, in your case along the cylinder the field will stay within the iron (a reasonable approximation) but at both ends they would escape and would need to loop around in a medium such as air or vacuum.
the second issue, as you mentioned is to get the current to correctly loop around, one way is to use the external boundary and make it conducting, but then the wires must extend to the boundary, with additional side effects, or you define the boundary along your cylinder ends, but this would avoid a natural looping of the B field.
Both approaches are approximations. Check some of the 3D coil examples in the model library
And for Eddy currents I believe you need to solve for V hence use MEF (check carefully with the doc) which is far heavier to handle and does not accept non conducting domains, so one must cheat a little and put a conduction on air 1-100[S/m] mostly does the job
--
Good luck
Ivar
I hope it will go on your workstation, but Eddy currents is far more trickier too set up and to solve.
One of the issues in 3D is that the magnetic field must loop around, that is that, in your case along the cylinder the field will stay within the iron (a reasonable approximation) but at both ends they would escape and would need to loop around in a medium such as air or vacuum.
the second issue, as you mentioned is to get the current to correctly loop around, one way is to use the external boundary and make it conducting, but then the wires must extend to the boundary, with additional side effects, or you define the boundary along your cylinder ends, but this would avoid a natural looping of the B field.
Both approaches are approximations. Check some of the 3D coil examples in the model library
And for Eddy currents I believe you need to solve for V hence use MEF (check carefully with the doc) which is far heavier to handle and does not accept non conducting domains, so one must cheat a little and put a conduction on air 1-100[S/m] mostly does the job
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi Ivar,
I have simplified the geometry of my model (induction motor) to reduce the degree of freedom and computational time. I have created edge on the indentity pair between the stator and the rotor. and free triangular mesh elsewhere.
I don't get any solutions. Everytime I simulate, I get the same error of singularity error, nonsymmetric matrix and like that.
I read in the discussion that poor mesh quality results to that error. Can you please check my mesh and suggest me what can I do now? Can you look at the error and suggest some remedy?
Thank you!
Regards
I have simplified the geometry of my model (induction motor) to reduce the degree of freedom and computational time. I have created edge on the indentity pair between the stator and the rotor. and free triangular mesh elsewhere.
I don't get any solutions. Everytime I simulate, I get the same error of singularity error, nonsymmetric matrix and like that.
I read in the discussion that poor mesh quality results to that error. Can you please check my mesh and suggest me what can I do now? Can you look at the error and suggest some remedy?
Thank you!
Regards
Attachments:
Please login with a confirmed email address before reporting spam
Posted:
9 years ago
Hi there,
Did you manage to find the problem with your model? I am facing a similar problem with my machine in 3D.
In my case, even though the solution converges, it does not make sense. That is, the flux from the stator does not seem to cross the airgap interface into the rotor, even though I have already applied the continuity condition on the identity pair at the boundary.
Any help in this regard would be appreciated.
Thanks,
Faisal.
Did you manage to find the problem with your model? I am facing a similar problem with my machine in 3D.
In my case, even though the solution converges, it does not make sense. That is, the flux from the stator does not seem to cross the airgap interface into the rotor, even though I have already applied the continuity condition on the identity pair at the boundary.
Any help in this regard would be appreciated.
Thanks,
Faisal.