CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   simple airflow through a box channel (http://www.cfd-online.com/Forums/fluent/46274-simple-airflow-through-box-channel.html)

Leo Lau October 9, 2007 09:56

simple airflow through a box channel
 
Hi, I am kinda new. I've modeled a simple airflow through a box, with all the walls as adiabatic walls and with a heat flux through the bottom wall into the airstream. As I would expect, the temperature of the bottom wall would increase rapidly in the beginning and will tends to reach an asymptotic value. The problem is that the temperature of the last node of the bottom seems to show some short of dip. So I would all the temperatures in a smooth line and then suddenly in the end, one point is below this line. Can somebody explain to me what is going on? Or what I am doing wrong?


Glenn October 10, 2007 06:42

Re: simple airflow through a box channel
 
I'm assuming from the above that the "last node of the bottom" is probably adjacent to an exit boundary of the solution. If so, you may be seeing a somewhat inaccurate exit boundary condition definition manifesting itself as a local pertebation (inaccuracy) in the solution. Look over the total solution and if all else seems correct, I'd ingore this last node result. If the last node is critical to your results, then perhaps revise the exit boundary condition to better approximate a more realistic exiting profile. You can write out the flow field solution on a plane close to the exit and map this result onto the exit boundary and iterate the overall solution further. Repeating this write/read a few times and you will probably have a sufficiently accurate exit boundary condition specified.

Leo Lau October 11, 2007 02:26

Re: simple airflow through a box channel
 
Thank you. The total solution of my problem seems to be correct, but I still find it strange that fluent does this. I also had a friend to simulate this in ansys cfx, which doesn't give this defect in the solution. I also have another question. Say I am using a symmetry plane in my simple airflow in the box channel. And this symmetry plane cuts my inlet into half. I divided my mass flow rate in the inlet by 2. I did this because I think I will get the same velocity profile. But I should I also divide my hydraulic diameter in the inlet boundary?

Glenn October 11, 2007 06:25

Re: simple airflow through a box channel
 
It sounds like Fluent gave you a localized flow anomoly while CFX did not. If the problem was locally ill-posed (ie: the exit boundary condition being imposed was locally inconsistent with the volume flow field approaching the exit)... Which solver gave you the more accurate result? How do you know? Please, don't asssume that because CFX did not give this anomoly, that it was therefore a more accurate result.

As to the symetry questions... I agree that for your problem the total inlet flow rate would be half, as you stated. However, I would not have divided the hydraulic diameter, as this parameter is also used in the various turbulence parameters, such as turbulent kinetic energy, intensity, dissipation. (See Fluent documentation). Altering the hydraulic diameter specified may well have inadvertently altered these parameters as well. Check your solution.

Hope this helps!

Leo Lau October 14, 2007 05:23

Re: simple airflow through a box channel
 
Both solvers gave me about the same answer, just a difference of less than 1%, only that fluent gave me that anomaly.' I guess you are right that I shouldn't have changed the value of the hydraulic diameter, it alters the turbulent kinetic energy in my simulation. Thank you.


All times are GMT -4. The time now is 04:33.