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 The information provided may be out of date.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 dependent heat flux (as a boundary heat source)

Please login with a confirmed email address before reporting spam

Hello all,

I am trying to model a 2D axi-symmetric analysis, which consists of a boundary heat source. I wish to apply a time-dependent heat flux for that boundary. The duration of the analysis is 30 days with 1 hour time steps (0,1[h],30[d]). The heat flux should be zero for hours 0-8 and 16-24; and q_boundary for hours 8-16 for each day.

I am setting the heat flux of the boundary heat source (General source, Qb) as follows:
Qb = q_boundary*(mod((t/3600),24)>=8)*(mod((t/3600),24)<16)

When I run the analysis, I see that no heat flux is applied at any time on that specified boundary. Could you please tell me where do I make a mistake?

Best regards,
--
Tolga

5 Replies Last Post 11.01.2012, 03:37 GMT-5
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 11.01.2012, 01:40 GMT-5
Hi

I suspect that your solver settings are such that it skips the time varying input flux (but it could be something else.

First of all, with HT I strongly recommend that you use a V4 Waveform "pulse" function (or a Heaviside function in 3.5), to smooth the transition, such steps gives very serious problems to the solver. Then you need to have a fine mesh on the heat source side (I often add a boundary mesh along the constant Q flux boundary, for time dependent studies) and finally you must help the solver with better steps. By default its in "free" stepping mode (far down in the time solver setting nodes), use rather " Intermediate" and then in your time series define time steps, at least 3 per rise and fall times, then longer in between.

The default time stepper settings is for settling i.e. exponential type asymptotic decays, and not good for periodic BC changes

--
Good luck
Ivar
Hi I suspect that your solver settings are such that it skips the time varying input flux (but it could be something else. First of all, with HT I strongly recommend that you use a V4 Waveform "pulse" function (or a Heaviside function in 3.5), to smooth the transition, such steps gives very serious problems to the solver. Then you need to have a fine mesh on the heat source side (I often add a boundary mesh along the constant Q flux boundary, for time dependent studies) and finally you must help the solver with better steps. By default its in "free" stepping mode (far down in the time solver setting nodes), use rather " Intermediate" and then in your time series define time steps, at least 3 per rise and fall times, then longer in between. The default time stepper settings is for settling i.e. exponential type asymptotic decays, and not good for periodic BC changes -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 11.01.2012, 02:57 GMT-5
Hi Ivar,

Thank you for your reply and comments.

I tried a different way after I sent the message. I constructed a time varying heat flux input function with a combination of rectangle and analytic functions.

Rectangle Function (rect1):
Lower Limit: 8
Upper Limit: 16
Size of transition zone: 2

Analytic function (q_boundary_operation):
Expression: q_boundary*rect1(mod(time/3600,24))
argument: time

Then, I used q_boundary_operation(t) for the boundary heat source. This way, I could be able to apply the time varying heat flux. However, the results were not as I expected, and I suspected something was wrong about time stepping. At that moment, I saw your reply, and when I changed the stepping mode to "Intermediate", the problem was solved.

However, in your message, there is a point which is not very clear to me, that is; you mentioned: "in your time series define time steps, at least 3 per rise and fall times, then longer in between". Is this the times we define in the time-dependent solver (i.e. range(0,1[h],7[d])) or is it in another node?

Thanks again,


--
Tolga
Hi Ivar, Thank you for your reply and comments. I tried a different way after I sent the message. I constructed a time varying heat flux input function with a combination of rectangle and analytic functions. Rectangle Function (rect1): Lower Limit: 8 Upper Limit: 16 Size of transition zone: 2 Analytic function (q_boundary_operation): Expression: q_boundary*rect1(mod(time/3600,24)) argument: time Then, I used q_boundary_operation(t) for the boundary heat source. This way, I could be able to apply the time varying heat flux. However, the results were not as I expected, and I suspected something was wrong about time stepping. At that moment, I saw your reply, and when I changed the stepping mode to "Intermediate", the problem was solved. However, in your message, there is a point which is not very clear to me, that is; you mentioned: "in your time series define time steps, at least 3 per rise and fall times, then longer in between". Is this the times we define in the time-dependent solver (i.e. range(0,1[h],7[d])) or is it in another node? Thanks again, -- Tolga

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 11.01.2012, 03:32 GMT-5
Hi

you have also the "Definition - Functions - Waveform - Pulse" for repeated pulses

My comment was jsut that you should ensure that your time stepping you define with your "range" has a few time steps during the transition, to help the solver identify the steps, otherwise it might miss some

--
Good luck
Ivar
Hi you have also the "Definition - Functions - Waveform - Pulse" for repeated pulses My comment was jsut that you should ensure that your time stepping you define with your "range" has a few time steps during the transition, to help the solver identify the steps, otherwise it might miss some -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 11.01.2012, 03:37 GMT-5
Thank you, Ivar.

--
Tolga
Thank you, Ivar. -- Tolga

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 11.01.2012, 03:37 GMT-5
Thank you, Ivar.

--
Tolga
Thank you, Ivar. -- Tolga

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.