Convection boundary conditions

[tusharve]

I am trying to set up a convection boundary on 3 surfaces of a rectangle 2D mesh.

What's the boundary condition syntax for that? I need conditions to be T(x,t)= hAdelT where delT is difference between surface temperature and ambient temperature.

Any help would be appreciated!

[tusharve]

I am trying to set up a convection boundary on 3 surfaces of a rectangle 2D mesh.

What's the boundary condition syntax for that? I need conditions to be T(x,t)= hAdelT where delT is difference between surface temperature and ambient temperature.

Any help would be appreciated!

For more explanation, I'm trying to set up a solid cylinder generating heat in contact with a 273K cold plate at the bottom and 296K temperature air on the other sides. Radial and axial conductivities are different

[david-zwicker]

Could please specify the exact mathematical formulation of your boundary condition. I don't know how to parse T(x,t)= hAdelT.

[tusharve]

Hi David, I'm setting up a convection boundary condition on 3 sides of a 2 D rectangular system and a cold plate conduction boundary on the 4th side.

So the conditions would be,

Q- heat flux
h- heat convection coefficient
delT- temperature difference between plate surface and ambient air
A-area of the plate (assuming unit thickness)

So the boundary condition should be,

Q=hAdelT

I have attached a copy of the boundary code I'm using currently.

I'm also trying to simulate the Matlab code here but with an axial component with a different thermal conductivity.

I appreciate the help. Thanks!

[david-zwicker]

I think you are very close already. You indeed need a mixed boundary condition for this case. py-pde expects the BC, to be of the form ∂c/∂n + A * c = B where the derivative is the normal derivative at the boundary. You thus need to bring you BC into this form. If you then have fixed coefficients A and B, you can simply use the boundary condition {'type': 'mixed', 'value': A, 'const': B}.

[tusharve]

Thanks David! I think this has solved my problem for the time being. I might develop a more detailed model in the near future.