spcify heat sink and temperature on a wall
 

Hi

I am trying to simulate conduction as seen in the attachment. I have a wall with as a heat source and another wall as a heat sink. However I want to define also a constant temperature on the heat sink. So the question is as follows:

- If I choose temperature boundary condition on a wall, does it mean the I allow a heat flux on this wall? Or do I need to specify for the solver that this wall is acually a heat sink?

thanks

Fluid
 

What's in-between the boundaries, a solid or static fluid or flowing fluid?

the whole body is a solid.

Conduction
 

If whole of it is solid then you don't need to use a tool. The temperature field will just be straight lines. And if you want to use tool, just apply heat on one side and temperature on the other.

Thanks for your reply Vinerem. The geometry which I want to simulate is much complexes, but I use the image in the attachment to discuss the boundary conditions.

I am doubtful if your suggestion would yield accurate results. Is it accurate to define a closed system and a heat source with no heat sink. ? I used some other software(Optistruct) to implement what you suggested and it never converges unless I define a heat sink.

Heat Source
 

Are you applying a heat source or a boundary condition (theoretically, both are same)? If you are apply heat source, then you will have to apply boundary conditions on both sides, top as well as bottom. If you are applying boundary conditions using heat flux and you apply only heat flux on both ends, then the solution is no unique. Assuming, you are apply heat source, it is the job of the thermal conservation equation to maintain the flow of the heat. But you will have to apply boundary conditions on top as well as bottom. If any one of those are missing, then the solution won't be unique. I don't know anything about optistruct but Fluent will assume the boundary to be adiabatic if no condition is applied.

Ok I guess the use of heat source became problematic here. So lets say I only use a positive heat flux on the upper side and a temperature on the lower side. My doubt is without the user defining a wall for the heat to exit from or a wall with negative heat flux, could I achieve accurate results or not.

I think if I define a positive heat flux on one wall, the I must define a negative heat flux on another wall to allow the heat exit from and keep energy balance and finally a temperature.

Temperature BC
 

The field variable for which the conservation of thermal energy is solved is T in case of constant properties. So, the specification of heat flux as a boundary condition essentially implies specifying gradient of temperature. And specifying T directly is used as it is. The flow of thermal energy is always from high to low thermal gradient. When heat flux is specified at one boundary, as per its direction, the heat flow either goes in or comes out. Due to this, a thermal gradient gets setup in the domain. When it encounters the temperature on the other boundary, the heat flow starts automatically depending upon whether the temperature on the other end is higher or lower than the temperature inside. So, don't worry. The conservation will be maintained properly with heat flux on one side and temperature on the other side.