CFD Online Discussion Forums - Mixed convection

Hello,

I'm interested in the Mixed-CHTC (Mixed Convective Heat coefficient)! Below you can read the problem description and other important information.

@CASE

c d -----------------------------' ' HOT_CEILING <--'e g h ' (318K) '-------' ' f ^ ' ' ' ' ' ' ' 'COLD_WALL Adiabatic ' ' (273K) ' ' ' ' ' ' ' ' Adiabatic ' -------------------------------------' a b

@Dimesions(in meters) with relative coordinates(X,Y):

a=(0,0) b=(2.78;0) c=(0;2.3) d=(2.28;2.3) e=(2.28;2.255) f=(2.28;2.2) g=(2.735;2.2) h=(2.78;2.2)

@ 1- T_exhaust = T_inlet (approximates the recirculation idea) It is a steady approach(in theory after certain of time) 2- HOT_CEILING had a uniform temperature distribution; this applies also for the COLD_WALL 3- a uniform velocity_inlet of 0,4 m/s into the cavity; so with a normal [x,y] of [-1,0] 4- a pressure_outlet with a Gauge Pressure = 0 and directional [x,y] = [0,1] 5- the velocity_inlet height=0,045 mm --> D(hydraulic)=0.084 6- Almost no turbulence at the inlet or the exhaust...so 5% turbulence intensity at the velocity_inlet. 7- AIR --> Compressible!!!? because the Brinkman-number>1 --> Br=2,25 --> (17,5*0,4^2)/(0,0276*45) 8- Boussinesq approximation for density is used;

@Solution Strategy:

FLUENT Version: 2d, dp, pbns, ske (2d, double precision, pressure-based, standard k-epsilon) Release: 6.3.26

Model Settings ---------------------------------------------------------------- Space:2d Time Steady Viscous Standard k-epsilon turbulence model Wall Treatment Enhanced Wall Treatment Heat Transfer Enabled Radiation None

Discretization Scheme:

Variable Scheme

------------------------------------------------

Pressure PRESTO!

Momentum Second Order Upwind

Turbulent Kinetic Energy Second Order Upwind

Turbulent Dissipation Rate Second Order Upwind

Energy Second Order Upwind

@PROBLEM:

No Converged results...very high instability of local temperature which i do monitor by looking at the bahaviour of the temperature at a certain point in the interior..For instance at (0,2;0,2).

I tried also:

1- decoupled solution: First get converged solution by setting Energy 'Off' in SOlution Control window...and indeed it did converged but after setting energy back 'on'.it begins te oscillate too much and so i get diverged results. Even by using UnderRelaxation Factor for Energy of 0,4 (Which is according to FluentGuide a problem to do so). The lowest factor might use is not lower than 0,8!

2- Step by step solution:

-I also tried to decrease the Rayleigh-number by power of 10. It seems also to converged. But after setting the gravity acceleration back to -9,81 m/s^2..it begint again to oscillate...and instabillity as result! - Also i tried a nother Solver --> The coupled implicit (or the Density Based Solver)...No convergence! - I tried also another turbulence models(KNG and k-omega), and other density approximations like incompressible ideal-gas...still no success. - I also changed the mesh..but it is still the same problem.

@QUESTION ARE THERE ANY OTHER IDEAS...What should be the problem? We have here a Forced convection at the ceiling and a natural convection at the left wall (COLD_WALL). Is there any other solution strategy, which should help me to getconverged results.

If you want to see the mesh, please contact me.

Thank you all,

Ehsan