transient conduction in a wall
 

this the problem. steam inlet from below the drum with temp 380 C, mass flow 0,4 kg/s with int condition 60 C (for wall and fluid), steam injected for 460 min. air inside the drum with temp. 60 C. here is the step what i've done.

gambit :
1.import model 2D
2. create face 1 for solid and create face 2 for liquid
3. mesh face for face 1 and face 2 together
4. create countinuum type face 1 solid and face 2 liquid selected
5. creat BC inlet at buttom line and symetry on left edge
6. export mesh

fluent 6 :
1. read case
2. scale
2. define model solver - unsteady
3. define mdl multiphase- eularian
4. define mdl energy - check
5. define mdl visc - k-epsilon
6. define material - vapor water (solid with rho=7850, cp =450, K = 40)
7. define phase- phase 1 (air) phase 2 (vapor)
8. define OC - gravity (-9.81)
9. define BC- inlet (phase 2, 0,4kg/s, temp 380 C) (phase 1, 0 kg/s, temp 60 C) there is wall 1 and wall 01:shadows appear at edge between solid and luqiud, BC without outlet

10. solve-control solution-ok
11. solve initialize- inlet
12. adapt region-insert number of x/y max/min - adapt - mark
13. solve initialize - patch (phase 2, liquid , value=0) (mixture, solid, value 60 C) this i want to make theres no vapor inside the drum and int temp. of wall is 60 C
14. solve monitor residual - plot
15. iterate - time step size (1) number of TS (76000 s) max iteration/TS (2)
16. iterate

there're all the steps friends, but the result temperature of the wall is not as i expected, its about 250 C and thats should be around 350 C. Please tell me what i missed. is that because i used fluent 6 ? i really need your help

thank you very much

do you have specific question or you want us to solve your problem

dear centurion2011

i just want to share my problem why i have different wall temperature from numerical result with measured result. i hope some one can explain to me. i write the steps because i thing there something wrong in my steps or something else. i am still learning using Fluent. sorry if my question too general or say something wrong in this forum

regards

maybe you should start with steady solution and then progress toward unsteady case.

For more info on unsteady problems, go to following link

Hope this will help you a bit :)

Since you used default solver settings. Check whether it is second order or higher. And don't use mesh adaption near boundary (if any), better to create a fine mesh than mesh adaption. k-epsilon turbulence model uses wall function which is not suitable for heat transfer applications try sst model.

Are you sure you are getting convergence with only 2 iterations per time step?

Daniele