CS inlet temperature

Hi,

I'm doing a test study of a rectangular enclosure with one wall as heat source (bottom), top (half only) as outlet and one side as the inlet(1/3 of the area). I put a negative velocity on the inlet to simulate a suction, thus direction of flow is actually from the "outlet" and going out thru the "inlet".
I added a -0.25 flux in the bottom wall as my heat source, inlet velocity of -0.25m/s and TempC=0. Initial value was set 25C (min:25C max:1000C).

At the end of my simulation I was surprised that the "inlet" temperature is hotter than the rest of the system.I think, the temperature of the "inlet" will depend on the temperature of the fluid passing it.
Please see attached and let me know if I did something wrong here.

Thanks

Florante

Post edited by: florante, at: 2010/08/11 13:36

Post edited by: florante, at: 2010/08/11 13:37<br /><br />Post edited by: florante, at: 2010/08/11 13:39

here is the attachment

thanks

Florante

Why don't you try instead a fixed value of temperature instead of impossing a flux? I'm also simulating an extracting hood system with one inyecting inlet (with hihger than ambient temp as fixed value), one suction inlet, an outlet and one of the walls has a fixed value temp much higher than the ambient (it would simulate the iron where you cook hamburgers). The resuls are what I expected.
You may try changing the sign of the flux too.

CAVT,

The objective of the test is to determine the temperature at some predefined points given some heating components like IC's or transformers where the flux is known. If I give the fixed value, can I solve for the flux with the component?

Thanks

Florante

Mmm... By fixing a value of temperature in a wall, inlet , etc., I think there would be no flux through that particular boundary, since a flux means that there's a change in the variable. I mean:

temp flux=0 , then temp=constant

That would be some kind of equilibrium condition for the boundary, but if you want to estimate the flux in a non-prescribed temp or flux area, I think you should be able to do it using the calculator filter of Paraview.
A note aside, the only reason I suggested setting fixed temperature values was only to check if the convection is being calculated properly (maybe I'm wrong, but I think a fixed temperature is mathematically easier and more stable to solve), not to make a final estimation.

What I still cannot figure out properly is why you set a negative heat flux, are you cooling that wall? If you're cooling the wall then a hotter suction inlet doesn't sound illogical. Try with a positive flux value if the wall is supposed to heat up, the flow direction and sense will finally determine the path of the convection. You may want to coarsen the mesh a bit or even try in 2D first to make fast calculations.

Hello,

Actually, fixing a temperature can lead to a heat flux. For example, if a fluid with inlet temperature 10° is flowing through a pipe, and you prescribe a wall temperature of 20°, it means you need to heat the pipe to maintain the wall temperature at the prescribed value, and that heat will be transferred to the fluid, hence a positive flux.

In Code_Saturne, the relation between the heat flux and the wall temperature depends on the turbulence model, so it is not straightforward (you would need to look at the CONDLI and especially CLPTUR subroutines to see how the Nusselt number is computed). For laminar flows, as there is no wall law, the exchange coefficient is simply based on the temperature gradient, and the heat flux is basically mass flow * conductivity * temperature gradient, but in most cases, you will be using a turbulent model.

This also means that the heat flux you obtain may be dependent on the turbulence model and wall law you are using.

Best regards,

Yvan