Discussion ClosedThis 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 transfer across a wall - issues with output results

Please login with a confirmed email address before reporting spam

Dear Community,

I have a problem with my simulation. Currently I am trying to solve a simple heat transfer problem across an externally insulated wall (you can find the file attached). In particolar, conduction inside the solid and convection at the surfaces. The problem is a time-dependent one, in which at the internal side I set 16 °C as air temperature and at the external side I applied an analytical function (sinusoidal). Initial condition inside the solid is 16 °C and convective coefficients are 7.69 and 25 W/m2 K. The simulation is performed with 1 min time-step and it is run for 10 days. The solver adopts the BDF method, and I decided to run a user-controlled tolerance simulation in which a value of 10^-4 and 10^-5 were set for the relative and absolute tolerance, respectively. The methodology I have selected was the manual one and not the factor method. The spatial discretization considered a default mesh, Normal.

The excpected results in terms of total heat fluxes calculated at the surfaces are sinusoidal functions of the time that oscillate around zero. Here it comes the problem, or better to say the problems.

What I get, in the specific, on the internal side, is a sinusoidal wave for the heat flux but not centered on the zero. It is like, i get a positive peak of 0.18 W/m2 and a negative one of -0.25 W/m2. If the magnitude of the heat flux had been higher, clearly this difference would have been negligible but here, since values of heat flux are low, this offset has an important impact. I've been trying to vary differently all the simulation parameters without any satisfactory result. This offset can be present but in percentage terms cannot be higher than 10% (better 7%).

Currently I changed the solver scheme into Runge-Kutta method, as suggested by the Manual for oscillatory problems but I am still waiting for results.

Do you have any suggestions? Has ever someone ecountered the same issue? I would like to understand more about settings of the simulation, e.g., relative and absolute tolerances, and so on. I read the manual but maybe information coming from practical examples can be more useful.

If I forgot some details, please tell me and I will reply.

Thanks for the help.



2 Replies Last Post 2023年11月28日 GMT+1 16:23

Please login with a confirmed email address before reporting spam

Posted: 10 months ago 2023年10月10日 GMT+2 17:03
Updated: 10 months ago 2023年10月10日 GMT+2 17:03

my comments as a newbie too.

I notice that your solver runs for 14,400s but your external temp period is 86,400s.

I suspect your results are due to simulating less than 1 day and getting the positive values of your Sine function.

try running for a full day.

Dan Field

my comments as a newbie too. I notice that your solver runs for 14,400s but your external temp period is 86,400s. I suspect your results are due to simulating less than 1 day and getting the positive values of your Sine function. try running for a full day. Dan Field

Please login with a confirmed email address before reporting spam

Posted: 9 months ago 2023年11月28日 GMT+1 16:23

my comments as a newbie too.

I notice that your solver runs for 14,400s but your external temp period is 86,400s.

I suspect your results are due to simulating less than 1 day and getting the positive values of your Sine function.

try running for a full day.

Dan Field

Thanks for the reply, I have just seen it now.

>my comments as a newbie too. > >I notice that your solver runs for 14,400s but your external temp period is 86,400s. > >I suspect your results are due to simulating less than 1 day and getting the positive values of your Sine function. > >try running for a full day. > >Dan Field Thanks for the reply, I have just seen it now.

Note that while COMSOL employees may participate in the discussion forum, COMSOL®software users who are on-subscription should submit their questions via theSupport Centerfor a more comprehensive response from the Technical Support team.

Baidu
map