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 The information provided may be out of date.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.

COMSOL help needed. (RF module - Heating of water using waveguide)

Please login with a confirmed email address before reporting spam

Hello there,
I am a relatively new COMSOL user who is in the process of validating a paper, to be able to get a better grasp at COMSOL.
The paper I am trying to validate using my simulation can be found here:
dspace.library.tu.ac.th/bitstream/handle/3517/1816/3/Waraporn_Experimental-1816.pdf?origin=publication_detail
On page 547.

I want to use a wave guide to send microwaves in the TE 10 mode to the water body in an air container.
The water body exhibits dielectric loss, The equations are found in table 1 page 550.

I have followed the microwave over(potato) model and feel like I've set up everything appropriately, however when I conduct the study the results indicate that the water body is not heating, perhaps there is a mistake in the way I defined the physics.
I've been trying to crack my head around this, but have had no luck.

In the attachment I've included my model. The paper also describes the boundary conditions in page 547.
The result I wish to achieve is the graph in page 553( Fig. 11 )

Any help or words on this matter will be greatly appreciated.

Kind regards,
Mohammed


13 Replies Last Post 28 mag 2014, 07:32 GMT-4

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 21 mag 2014, 11:42 GMT-4
In the model that you have uploaded you have not defined the imaginary part of the permittivity of the water. This is exactly the parameter that defines the losses.
In the model that you have uploaded you have not defined the imaginary part of the permittivity of the water. This is exactly the parameter that defines the losses.

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 21 mag 2014, 12:02 GMT-4
But I have defined the real part and the loss tangent(delta)
Shouldn't that be enough?
But I have defined the real part and the loss tangent(delta) Shouldn't that be enough?

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 22 mag 2014, 02:59 GMT-4
It is enough information because the imaginary part can indeed be calculated from the real part, the loss tangent and the conductivity (en.wikipedia.org/wiki/Dielectric_loss). But maybe COMSOL does not do this calculation and you have to do it yourself?
It is enough information because the imaginary part can indeed be calculated from the real part, the loss tangent and the conductivity (http://en.wikipedia.org/wiki/Dielectric_loss). But maybe COMSOL does not do this calculation and you have to do it yourself?

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 22 mag 2014, 09:23 GMT-4
I took your advice Pieter and included the imaginary part too.. which is simple tan(delta) * real part.
However there is no difference in the results obtained. The water body seems to be not heating up at all.
I took your advice Pieter and included the imaginary part too.. which is simple tan(delta) * real part. However there is no difference in the results obtained. The water body seems to be not heating up at all.

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 22 mag 2014, 09:32 GMT-4
I see. I cannot evaluate your model myself, i get the error

Nonlinear solver did not converge.
In segregated group 1:
Time : 10.02836128938059
Segregated group 1

Undefined value found.
Undefined value found in the stiffness matrix.
For mesh-case 1 there are 316 equations giving NaN/Inf in the matrix rows for the variable mod1.E10.
and similarly for the degrees of freedom, NaN/Inf in the matrix columns.
Last time step is not converged.
I see. I cannot evaluate your model myself, i get the error Nonlinear solver did not converge. In segregated group 1: Time : 10.02836128938059 Segregated group 1 Undefined value found. Undefined value found in the stiffness matrix. For mesh-case 1 there are 316 equations giving NaN/Inf in the matrix rows for the variable mod1.E10. and similarly for the degrees of freedom, NaN/Inf in the matrix columns. Last time step is not converged.

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 22 mag 2014, 09:34 GMT-4
I included in the attachment a solution to the error you are getting.
This should not give you any errors.

However the results as you can see if you go to study are incorrect.
I included in the attachment a solution to the error you are getting. This should not give you any errors. However the results as you can see if you go to study are incorrect.


Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 22 mag 2014, 11:04 GMT-4
You have no electric field inside your water. Maybe you should include the water domain (2) in the 'wave equation, electric 1' node.
You have no electric field inside your water. Maybe you should include the water domain (2) in the 'wave equation, electric 1' node.

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 22 mag 2014, 11:33 GMT-4
I am not so sure about that Pieter. Because The domain for the water body is assigned to Microwave heating model. THe wave equation, electric is only there to show that the air domains need not take part in the microwave heating physics.
I am not so sure about that Pieter. Because The domain for the water body is assigned to Microwave heating model. THe wave equation, electric is only there to show that the air domains need not take part in the microwave heating physics.

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 23 mag 2014, 02:39 GMT-4
Well, what I see in the potato heating example from the model library is that the electric field node is defined also in the material that is heated. Subsequently the electromagnetic heat source is only defined in the potato so that only there the heating happens.
Well, what I see in the potato heating example from the model library is that the electric field node is defined also in the material that is heated. Subsequently the electromagnetic heat source is only defined in the potato so that only there the heating happens.

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 23 mag 2014, 15:01 GMT-4
Yes that's correct Pieter.

I tried to use the rf waveguide model, to try and simulate this situation.
Everything runs and simulation converges now, however the results are impossible..
in 60 seconds the maximum temperature reached is 120 degC..
Perhaps there is something wrong with the physics again.

Another thing that I noted is if I change from loss tangent to relative permittivity in the microwave model physics, my solution does not converge(doesnt give me an error but keeps runnign, I left it on for 2 nights..)

Any help would be greatly appreciated.


Edit: I have included in the attachment the new model.
Yes that's correct Pieter. I tried to use the rf waveguide model, to try and simulate this situation. Everything runs and simulation converges now, however the results are impossible.. in 60 seconds the maximum temperature reached is 120 degC.. Perhaps there is something wrong with the physics again. Another thing that I noted is if I change from loss tangent to relative permittivity in the microwave model physics, my solution does not converge(doesnt give me an error but keeps runnign, I left it on for 2 nights..) Any help would be greatly appreciated. Edit: I have included in the attachment the new model.


Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 26 mag 2014, 03:35 GMT-4
Yes this indeed strange, even if all the energy inserted at Port 1 would go into the water, 60 s at 300 Watt should only increase the temperature of this amount of water by something like 35 K.

What is the purpose of Port 2?
Yes this indeed strange, even if all the energy inserted at Port 1 would go into the water, 60 s at 300 Watt should only increase the temperature of this amount of water by something like 35 K. What is the purpose of Port 2?

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 26 mag 2014, 10:38 GMT-4
Port 2 is unexcited, so it would be a way to absorb the incoming microwaves from port 1 I believe.
Port 2 is unexcited, so it would be a way to absorb the incoming microwaves from port 1 I believe.

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 28 mag 2014, 07:32 GMT-4
Ok, so that is what would happen if you would hack your microwave oven to work with the door open. But that is no explanation of why your water heats so fast, I would instead expect a lot of energy to escape through port 2, so that less of it ends up in your water. Frankly I have no idea of why your results are not correct. Have you tried if mesh refinement/coarsening has any influence on the result?
Ok, so that is what would happen if you would hack your microwave oven to work with the door open. But that is no explanation of why your water heats so fast, I would instead expect a lot of energy to escape through port 2, so that less of it ends up in your water. Frankly I have no idea of why your results are not correct. Have you tried if mesh refinement/coarsening has any influence on the result?

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.