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.

temp. Step function

Please login with a confirmed email address before reporting spam

hi there,

I would like to input the boundary temperature of my model as step function such as

T1= 300K T2= 350K at 1000 sec.

I tried to do by create the step function but I failed to use it as boundary ??

so can anyone help me with it?

thanks
Abdul

10 Replies Last Post Aug 22, 2011, 2:30 p.m. EDT
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 Aug 8, 2010, 1:40 a.m. EDT
Hi

are you talking about a time step function ? and are you in V3.5a or v4 ?
the way to write out the equations are slightly different,
as time function somthing like but it could then be done T1+step(t-time0,step_rise)*(T2-T1)
where time0 is the moment of the step and "step_rise" is the half rise time

step does not exist as such in V3.5a its called flc1hs see command.pdf

--
Good luck
Ivar
Hi are you talking about a time step function ? and are you in V3.5a or v4 ? the way to write out the equations are slightly different, as time function somthing like but it could then be done T1+step(t-time0,step_rise)*(T2-T1) where time0 is the moment of the step and "step_rise" is the half rise time step does not exist as such in V3.5a its called flc1hs see command.pdf -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Aug 8, 2010, 7:24 p.m. EDT
Yes, I'm talking about time step function on V4.

And I would like to simulate the temperature around an enclosure to be variable between T1 and T2 for 10000 sec. .And the boundary temp. Rise or fall at every 1000sec.

How to write the equation for boundary Temp.?

where I could find the command.pdf?

thanks for you help,
Abdul
Yes, I'm talking about time step function on V4. And I would like to simulate the temperature around an enclosure to be variable between T1 and T2 for 10000 sec. .And the boundary temp. Rise or fall at every 1000sec. How to write the equation for boundary Temp.? where I could find the command.pdf? thanks for you help, Abdul

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Sep 30, 2010, 6:38 a.m. EDT
Just search for "command.pdf" and you'll find it on your computer...
Just search for "command.pdf" and you'll find it on your computer...

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Oct 16, 2010, 9:15 a.m. EDT
Hi,

I also seeking for advice on the temp step function.
I trying to heat up my model from 300K to 500K at the first 360s and then turn off the heat and cool down to room temp. The module I use is Structural Mechanics/Thermal stress(Time dependent). Iuse the heat source under this module with 10W/m3 * (t<370). The time setting is range (0,10,24hrs). I thought after 360s, the heating will turn off and it will cool down to room temp but I was wrong.
Could you please advice how I can simulate my model by anneal up to 500K for 360s and then let it cool down to room temp? Should I use step function in this case?
look forward to the advice and suggestion.
Thanks
Hi, I also seeking for advice on the temp step function. I trying to heat up my model from 300K to 500K at the first 360s and then turn off the heat and cool down to room temp. The module I use is Structural Mechanics/Thermal stress(Time dependent). Iuse the heat source under this module with 10W/m3 * (t

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 Oct 16, 2010, 10:55 a.m. EDT
Hi

I can only advise to considert to use the Heavyside functions instead for a direct step (search the forum and your doc), and force some smaller time steps around the transitions that should help for the convergence.

I would also check the results with a strict time stepping, to be sure the automatic solver dooes not miss the turnoff time by too far

--
Good luck
Ivar
Hi I can only advise to considert to use the Heavyside functions instead for a direct step (search the forum and your doc), and force some smaller time steps around the transitions that should help for the convergence. I would also check the results with a strict time stepping, to be sure the automatic solver dooes not miss the turnoff time by too far -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Oct 16, 2010, 3:05 p.m. EDT
Do you mean I can use the heaviside function to setup the annealing process like hea the model from 300K to 500K for the first 360s and then coold it down to room Temp? Is it another type of step function allow turn on and off with specific time/temp??
Could you tell me where to find the explaination of how to use the heaviside function?I couldn't find it in comsol.
I'm using comsol 4a. didn't see the heavifunction under the Global Definitions.
Thanks!!

Do you mean I can use the heaviside function to setup the annealing process like hea the model from 300K to 500K for the first 360s and then coold it down to room Temp? Is it another type of step function allow turn on and off with specific time/temp?? Could you tell me where to find the explaination of how to use the heaviside function?I couldn't find it in comsol. I'm using comsol 4a. didn't see the heavifunction under the Global Definitions. Thanks!!

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 Oct 17, 2010, 4:20 a.m. EDT
Hi

the doc of 4 is not complete yet, there is more in the 3.5a doc.

But if you are in V4, then the step function are ALREADY including Heaviside functions for smoothings so it's just to be sure you master all parameters (also those in the sub-fields not necessarily open by default), always plot and check the functions you create (and make "functions" and analyse them separately before calling the functions in the BC fields, rather than putting the equations directly in the BC fied.

I was mainly pointing out that in time series solving, non-continuous step functions give often trouble to the solvers, so be sure you have some smoothened contnuous functions, and that, when you know the transition times, you give specific time steps with suffiecient resolution around these transitions.
Else the solvers might play you a game and give false results, even when converging correctly

--
Good luck
Ivar
Hi the doc of 4 is not complete yet, there is more in the 3.5a doc. But if you are in V4, then the step function are ALREADY including Heaviside functions for smoothings so it's just to be sure you master all parameters (also those in the sub-fields not necessarily open by default), always plot and check the functions you create (and make "functions" and analyse them separately before calling the functions in the BC fields, rather than putting the equations directly in the BC fied. I was mainly pointing out that in time series solving, non-continuous step functions give often trouble to the solvers, so be sure you have some smoothened contnuous functions, and that, when you know the transition times, you give specific time steps with suffiecient resolution around these transitions. Else the solvers might play you a game and give false results, even when converging correctly -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Nov 22, 2010, 11:41 a.m. EST
Hi,

I have a question on the Postprocessing. My simulation looking for Hydrostatic stress and its gradient, but I couldn't find it in the Plot Parameter/ Predefined quantities. The hydrostatic stress is the average of 3 normal stress(x,y and z). These normal stresses are provided in the predefined quatities after simulation. Could you advice or provide the step how I can define the hydrostatic stress + gradient so that the result will plot on my model(surface,subdomain...)?

Thanks
Hi, I have a question on the Postprocessing. My simulation looking for Hydrostatic stress and its gradient, but I couldn't find it in the Plot Parameter/ Predefined quantities. The hydrostatic stress is the average of 3 normal stress(x,y and z). These normal stresses are provided in the predefined quatities after simulation. Could you advice or provide the step how I can define the hydrostatic stress + gradient so that the result will plot on my model(surface,subdomain...)? Thanks

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 Nov 22, 2010, 2:45 p.m. EST
Hi

unfortunately Ido not know it that variable is defined within COMSOL. One way around is to activate the equations sub-node (options preferences ...) and then scan the variables in your main physics.

if not, and you know the formula you can make a "user defined Variable" and equate it to your stress value, then its available over the Entities you select (or globally)

--
Good luck
Ivar
Hi unfortunately Ido not know it that variable is defined within COMSOL. One way around is to activate the equations sub-node (options preferences ...) and then scan the variables in your main physics. if not, and you know the formula you can make a "user defined Variable" and equate it to your stress value, then its available over the Entities you select (or globally) -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Aug 22, 2011, 2:30 p.m. EDT
sir,
my problem is that..i have to varry my temp from 0 to 1000k continuosly and shud analyse d fatigue nature created..so how can i use the step function here..i dint get d way of aproach..i made a step function called 'step1'..later gow can i use this function in place of the variable'temperature)..?
sir, my problem is that..i have to varry my temp from 0 to 1000k continuosly and shud analyse d fatigue nature created..so how can i use the step function here..i dint get d way of aproach..i made a step function called 'step1'..later gow can i use this function in place of the variable'temperature)..?

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.