Made some modifications to step26, but the result is incorrect

[Tiny Y]

Hi everyone,

I have made some changes to step26:
1. took theta as 1.
2. multiplied the coefficients cd(volumetric heat capacity ) and kt(heat transfer coefficient) in front of mass matrix and stiffness matrix to 10^6(J/(m^3·℃)) and 0.34（W/(m·℃)) respectively.
3. imposed Dirichlet boundary conditions (colorize=true, -50 ℃ for boundary id=0, 0 for boundary id=2,4, and took the original periodic RHS of step-26 to 0)

Then I noticed that positive temperatures appear in the interior of the domain.
What's the cause of this?
Any help is appreciate.

[Wolfgang Bangerth]

On 8/4/24 02:52, Tiny Y wrote:
> :

> 1. took theta as 1.
> 2. multiplied the coefficients cd(volumetric heat capacity ) and kt(heat
> transfer coefficient) in front of mass matrix and stiffness matrix to
> 10^6(J/(m^3·℃)) and 0.34（W/(m·℃)) respectively.
> 3. imposed Dirichlet boundary conditions (colorize=true, -50 ℃ for boundary
> id=0, 0 for boundary id=2,4, and took the original periodic RHS of step-26 to 0)

> 微信图片_20240804165104.png
> Then I noticed that positive temperatures appear in the interior of the domain.
> What's the cause of this?

How do you choose the time step? If you change the coefficients, you in
essence change the *time scales* at which change happens. You have to adjust
the length of the time step accordingly.

Best
W.

--
------------------------------------------------------------------------
Wolfgang Bangerth email: bang...@colostate.edu
www: http://www.math.colostate.edu/~bangerth/

[Tiny Y]

HI,

Thank you for your reply!

 I found that by setting the time step large enough to around 10^3 the results look correct!

Best 

Y.