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.

Joule Heating - Resistive Heaters

Please login with a confirmed email address before reporting spam

Hi,

I've built a COMSOL model that models the joule heating of electrodes that are designed to heat up channels in a microfluidic device. I am able to choose a V0 (input voltage) run a transient study and get back temp values very close to what I expect so I think my model is sound.

What I want to do now is vary the voltage applied over time. Lets say some type of square function that applies 5V for 2 second 2V for 2 sec and see how that will affect the heat in my system over time. When I try using the function: V0 = 5V*(sin(2*pi*t/4)>0) + 2V*(sin(2*pi*t/4)<=0) to define the input voltage my model does not solve giving me the following error: "Repeated error test failures. May have reached singularity at time: 0. Last time step is not converged." I tried this first because I have used this technique before with a boundary temperature source to apply a varying temperature.

This is obviously not the best way to vary the voltage within the model. Is there a better way to do this? Would it be some kind of periodic condition set on the boundary heat source? I am new to COMSOL so detailed explanations would be greatly appreciated.

Thanks

3 Replies Last Post 19 sept. 2013, 15:00 UTC−4

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 18 sept. 2013, 03:41 UTC−4
Hi Alex,

I think Comsol is right, there is a singlarity at t=0. The way you have defined
the voltage that are jumps.

I suggest to use a smooth function. The "step" function in Comsol allows for
smooth step-function. The step can be very fast, but smooth is better.

Also from my experience it is better not to start a transient simulation at the
time of such an event, start shortly before.

Regards

Jens
Hi Alex, I think Comsol is right, there is a singlarity at t=0. The way you have defined the voltage that are jumps. I suggest to use a smooth function. The "step" function in Comsol allows for smooth step-function. The step can be very fast, but smooth is better. Also from my experience it is better not to start a transient simulation at the time of such an event, start shortly before. Regards Jens

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 19 sept. 2013, 11:46 UTC−4
Thanks Jens,

I am trying to build an analytical function that combines 4 steps to define my voltage cycling. However I can't figure out how to normalize the function. The steps go from 0 - 1, 1 - 0.4 , 0.4 - 0.7, and then 0.7 - 1.0, and when I add them it normalizes 2.1 - 3.1. How can I normalize this back to 0 - 1 on the y-axis?

Also, I noticed that you can add a periodic extension to an analytical function. Does this mean that I can make the function repeat itself because ultimately I want this cycle to repeat itself several times. The help in COMSOL doesn't really help and the parameters don't really make sense. Does anyone have any insight on this feature?

Thanks,
Thanks Jens, I am trying to build an analytical function that combines 4 steps to define my voltage cycling. However I can't figure out how to normalize the function. The steps go from 0 - 1, 1 - 0.4 , 0.4 - 0.7, and then 0.7 - 1.0, and when I add them it normalizes 2.1 - 3.1. How can I normalize this back to 0 - 1 on the y-axis? Also, I noticed that you can add a periodic extension to an analytical function. Does this mean that I can make the function repeat itself because ultimately I want this cycle to repeat itself several times. The help in COMSOL doesn't really help and the parameters don't really make sense. Does anyone have any insight on this feature? Thanks,


Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 19 sept. 2013, 15:00 UTC−4
I solved the normalization issue. But I am still curious about the periodic function.
I solved the normalization issue. But I am still curious about the periodic function.

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.