dear friends

my problem is the "shell conduction" function of the fluent. i couldn't understand the use of this function. and wonder if i have to use it on my project. i'm working on natural convection between horizontal parallel plates where parallel plates have a constant plate and open to the surrounding. am i going to use shell conduction for the wall boundaries of my problem?

Shell conduction is an extension of the thin-wall or wall-shwadow:wall concept. Since this poor post message panel does not allow ascii drawings i will refer to picture 7.13.2 in the fluent manual for further explanations. Lets add another temperature Ta on the left side of this wall (where those fluid/solid cells are) and if Ta is not equal Tb heat flux through the wall will occur (in direction of x). Now lets add some more temperatures to the picture, Tc at the bottom of the wall and Td at the top. If Tc is not equal Td heat flux within the wall will occur. Without shell conduction activated fluent will not compute this heat flux. With shell conduction fluent will try to solve the heat flux within this "thin wall". This leads in my experience to poor convergeance and if you are really interested in the temperature profile of that wall you should spend some cells on it.

Hope this helps

RoM

firstly thx for response!

as i mentioned i 'm working on "effect of plate seperation on two known sized horizontal parallel plates where bottom plate is hotter than the top plate"

my boundary condition on walls (plates) is fixed temperature not heat flux. and i'm looking for surface nusselt numbers on the bottom plate. if i use shell conduction function the nusselt number values became smaller than the solution without shell conduction function.

so i couldn't understand which is true for my situation and boundary condition. am i going to use shell conduction function or not?

hope to hear some advice. thx a lot!

Since it hard to say "use shell conduction or or not" the best option are some experimental calculations. In your case, i would remove the top plate and calculate only the free convection from the hot bottom plate. This is a standard case and many experimenal datas aswell as empirical formulations for Nusselt numbers are available. Do some calculations with and without shell conduction and try all those k-e turbulence modells (standard k-e is not very sensitive to natural convection). Compare your results to the experimental datas and choose the the setup wich will fit best. Then you can add the top plate and continue the calculation.

Hope this helped

RoM

[shahid.araman]

hii
im working on fin heat transfer. i need to know whether this shell conduction will help me.