Discussion Forum

I actually just want to know the temperature after 1h in the coil winding.

My questions are:
1. Which physics are most suited? (Now I have Magnetic Field and Heat Transfer Solids)
2. Heat Source: If I use Magnetic Field and Heat Transfer Solids physics, should I choose Electromagnetic heating as heat source, or should I define it by myself as a value [W/m^3]. If I define it by myself how do I know the power?
3. Is Time Dependent Study the best alternative?
4. For result I take Line integral around the conductor.
5. I use Single Turn Coil to apply current to the coil winding.

The problem is that the temperature just rise and rise and becomes very high, without stabilizing. It is like there is no air to transport the heat away.

Sorry if my problem is not very clear.
Thanks

Hi

You should ask what is your heating and heat exchange process ?
What I understand from your questions you have Joule heating from the current in the coil, and then conduction in the wire (this probably could be ignored if the wire section is small compared to the overall dimensions of your model) and the air convection. In that case it's the "Joule heating" module or the EC+HT combination that is the most appropriate, the EC defines the current flow, the material resistance and the current gives the heat generated and the surfaces exposed to air gives the loss via a heat boundary convection loss (typically some 5[W/(K*m^2)] but could be also 2-10 for non forced heat convection in air). By choosing two physics that interact in the recent versions you must ensure that your physics are correctly coupled under the "multiphysics tab".

There are many similar examples in the model library, take a closer look ;)

--
Good luck
Ivar

Thank you

I am sorry I still have some question after implementing Joule Heating.

1. My winding model is an array of conductors, total 96 conductors. With air between them. I add a current source and put the current density (A/m^3) for one conductor. Because the volume of one conductor is very small and the current is 400A the current source becomes huge, about 1 billion A/m^3.

2. As heat source is it okay to choose the "Total power dissipation density"? Or should I put something else?

3. When adding Joule Heating 3 subgroups of Multiphysics appear.
- Electromagnetic Heat Source
- Boundary Electromagnetic Heat Source
- Temperature Coupling
Do I have to change anything here?

( Everything is Time Dependent)

When I run I get Error "Failed to find consistent initial values. Last time step is not converged."
I don't quite understand.

Thank you for helping

Hi

You have many questions, and COMSOL implies many many settings, too many to cover all here :)

So I would suggest that you start simpler, make a very simple model (i.e. 2D axi one coil section or take a few of the Application library cases and study these in details) and add the physics one by one and check the settings you need

For your error message that one comes mostly when you have initial conditions that are too far from the solution. Often its worth to run a first stationary case to start a time series in a known logical initial condition, but again it all depends on your model and values you have

--
Good luck
Ivar

Hello, thank your reply.

My model is already very simple. The file is attached if you want to have a look.

Now I have the problem that the heat rise in a conductor is linear and does not seem to come to a steady state. I need some natural air convection or something that describes air flow around the conductors.

Thank you for reply