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 problem solution doesn't depend on time
Posted Sep 21, 2011, 12:17 p.m. EDT 10 Replies
Please login with a confirmed email address before reporting spam
I solve heat transfer, flow and PDE projects together. My problem is time-dependent. Comsol successfully solves it but there is a problem with the solution. When I plot 2D temperature distribution and start changing the time, the value of the temperature is not changed. I have read in the forum that maybe this problem comes from initial values but I can't come up with a solution for it.
Please help. Any advice is highly appreciated.
Thanks in advance.
Please login with a confirmed email address before reporting spam
re you sure you are adding some heat energy (flux) to your model, thatyou have selected a BC node WITH a boundary defined. Mostly when the temperature is not changing, it means that you have no energy input (heat flux on) to your system
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
I have added a Heat Source in the Heat Transfer physics. My heat source is a Gaussian beam.
Q = (2*alpha*Po/(pi*r0^2))*exp(-2*(r/r0)^2)
I maintain temperature fixed on the walls of the cell using Temperature (T0=293K). (I have done this because the cell is placed in the thermostat). When I remove it, temperature becomes extremely high.
Velocity is also constant when changing the time.
p.s. I get a distribution of the temperature but it doesn't change with time. Temperature distribution is like a Gaussian distribution. find attached.
From the experiment I know that temperature should gradually rise then it becomes constant after several seconds.
Attachments:
Please login with a confirmed email address before reporting spam
then I missed something ;)
But you need to define a boundary with a constant temperature (or another heat sink) A constant T boundary means that COMSOl adapts the heat flux through this boundary to keep the Temperature constant hence to get an energy flux balanced. if no you have a perfectly isolated sytem, and then if you heat it , with time you will end up at a T=inf ;)
You can also hep debugging your model by plotting the heat fluxes or total energy (but note that you might have quite some discrepancies on heat flux integrations. See the doc
Another point, if you have solved the model a few times and updated it, COMSOL might have generated several Study DataBases and you are updating no 2 or 3 while your plot is pointing to the study 1
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
I want to obtain some information about convection induced by Gaussian laser beam. I use two physics, laminar flow and heat transfer. In laminar flow physics I use initial values
velocity = 0 and pressure = rho*g_const*(height of the cell - z)
and
volume force = spf.rho*g_const*betta*(T-T0) , betta = 0.002[1/K]
In heat transfer i just fixed temperature on the walls (T0=293K) and added above-mentioned Gaussian heat source.
Initial temperature is T0.
Please login with a confirmed email address before reporting spam
Where do you place the heat source? If your walls have constant temperature it is no surprise T does not vary with time, then it all depends on where you placed your source and how strong it is.
Cheers
Please login with a confirmed email address before reporting spam
Thank you for response.
I have fixed the temperature on the boundaries, because without that point temperature becomes extremely high.
I also tried to solve this by means of "Inward heat flux", using air thermal conductivity=10 W/m^2/K, but it didn't help. Temperature was so high.
Do you have any suggestions? There is also one point always bothering be. I set initial value for temperature T0=293.5K. After calculation I get T(t=0) = 295K. The same can be said of velocity. I set v(t=0)=0, but after calculation it is about 10^-9. ???
Any help or thought is highly appreciated.
Please login with a confirmed email address before reporting spam
Ozgur
Hi Amir,
Thank you for response.
I have fixed the temperature on the boundaries, because without that point temperature becomes extremely high.
I also tried to solve this by means of "Inward heat flux", using air thermal conductivity=10 W/m^2/K, but it didn't help. Temperature was so high.
Do you have any suggestions? There is also one point always bothering be. I set initial value for temperature T0=293.5K. After calculation I get T(t=0) = 295K. The same can be said of velocity. I set v(t=0)=0, but after calculation it is about 10^-9. ???
Any help or thought is highly appreciated.
Please login with a confirmed email address before reporting spam
The fact that T gets extremely high if you do not constrain the wall T to a fixed amount seems to point out the fact that you may not set a constant T if you want to see it change over time.
10^-9 is zero numerically, just a little moving around zero. 295 is harder to explain, but it seems you have non consistent BCs.
As has been pointed out, what is your time scale (but I don't think it is critical since you do see change over time if you don't constrain teh wallT)? What are dimensions of your geometry? What is the material?
Cheers
Please login with a confirmed email address before reporting spam
At first, thank you for your good advices.
I have read in the forum that Comsol changes initial values to find a convergent solution for the problem. I think that's the problem why I get other values after solving the problem. But in my problem values of variables are small. That's why I am concerned with the initial value of the velocity after solving the problem. Expected value of the velocity is about 15 um/sec but I get 10^-9 instead. My domain height = 100um=0.0001m and width = 200um, but the with is not playing any role here.
In Heat transfer as an initial value T0=293.5K has been used.
In Laminar flow Velocity(t=0)=0 (both r and z components).
In laminar flow physics I don't have any boundary conditions. There is no such option in this physics as far as I know. In Heat transfer physics I just fixed temperature on the walls.
About time scale. From the experiment it was ascertained that I can solve this problem within 10sec. After 10sec. material becomes isotropic and my calculations become meaningless.
About constant temperature on the walls. I tried to remove it and now I see the evolution of the temperature but It becomes extremely high within the first second. As always, temperature is extremely high - more than 600K.
Thanks
Please login with a confirmed email address before reporting spam
It would be better to have your model, but from what you said it seems that velocity and T are calculated all right; if the values do not seem to be correct it may actually be an issue with the intensity of your heat source (too high maybe) and with your fluid dynamics boundary conditions.
What is the value of all teh parameters in your heat source? What is your inlet? And the outlet?
Cheers
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.
Suggested Content
- FORUM Two solver steps in time depend. study
- KNOWLEDGE BASE Improving Convergence of Transient Models
- BLOG How to Save Computational Time with a One-Way Coupling Approach
- KNOWLEDGE BASE Solving electromagnetic heating models with high-frequency loads
- FORUM Solution to time depend problem doesn't change with respect to time.