[dlalstnf]

Hello, everyone.

I run SU2_CFD solver with isothermal boundary condition(Twall=300K).

Governing Eq. is RANS (SA model).

I think that the residual calculation on the boundary wall (isothermal calc.) is implemented correctly, but I cannot find out primitive variable updating part..

I think temperature (primitive variable) is not fixed as a Twall(input value).

The problem is that the temperature distribution is not 300K in the result file (.vtu, paraview) rather there is a variation in temperature in the result file.

Is there anyone who knows about this problem??

[bigfootedrockmidget]

temperature is imposed as a weak boundary condition, so you only get the imposed boundary value in the limit of complete convergence.

[dlalstnf]

Thank you for your kind reply!

I didn't know about "weak" boundary condition.

Then, let me ask one more question,

Is heat flux correct as a result of the iso_thermal boundary condition?
I mean, when I visualized the intermediate result file (.vtu), is heat flux have a trust worth?

[dlalstnf]

Dear, bigfootedrockmidget

Could you explain more about the weak boundary condition in SU2?
or Do you know any paper to which SU2 referred?

Thank you!

[pcg]

The way it works is by imposing a heat flux based on the temperature difference between the imposed wall temperature and the current temperature in the interior of the domain (closest normal point from the wall).
This ensures global conservation of energy. Forcing temperature in a strong way is not conservative (energy in != energy out until the mesh is very fine).

[dlalstnf]

Could you give me a reference in detail? I want to know how the isothermal boundary condition is implemented in SU2 theoretically.

[bigfootedrockmidget]

If you search for weak vs strong boundary conditions for finite volume methods, you can find some helpful links. It is mentioned for instance in this Stanford course, slide 22-23:

https://web.stanford.edu/group/frg/c...-AA214-Ch8.pdf