Hi everyone
I have been doing an analysis to simulate an experimental setup using ansys fluent. It basically consists of a shell-and-tube heat exchanger. I have designed my model and created the mesh. When I run the calculations I get 292.2181K for the hot fluid outlet temperature, while the experimental temp was 292.8841K, which seems to be an small and acceptable difference. However, when I analyze pressure drop, ansys simulation gives 0.09 kPa/m while the experimental result is 0.3 kPa/m. So, could you write your comments about this difference on pressure drop and if there is any way to improve it? Currently, I used k-e stander wall functions because it is the model that yields closer temperature results and for method I used SIMPLE. Also, hexahedral mesh gave closer temp results when compared to tetrahedral mesh. Thank you all in advance.
I suggest that you check the dimensions of the geometry and all input numbers of all variables including velocity or flow rate, density and viscosity (they both need to be functions of temperature), etc.
After the above check, you need to know the actual roughness of the walls of the experimental results and then, in FLUENT you may input that roughness value. However, mesh requirements will change when you decide to consider the wall roughness.
@Ahmad Khalaf Sleiti, thanks for your suggestions. I'll take a look at that roughness parameter because I did not really think of that before calculation. Regards, Humberto
Yes. In actual condition wall friction (due to roughness) drops the pressure. So you should measure the roughness parameter and put it in ANSYS. I am agree with Dr. Ahmad Khalaf Sleiti .
I would like to thank all of you for sharing your thoughts. I did some research based on Ahmed comment and I could gather the data needed to compute the roughness of the tube surface. It is working as expected now.
Dear santos, first you need to make sure your mesh is stable regardless its type. if you could get good mesh qualities with the automated options that will be the best one "fast and most accurate".
Next, you need to identify the best numerical model, if you in rush start with k-e relaizable, scalable wall function.
finally, don't compare dP, compare friction factor instead as it will not be affected by the dimensions.
wish that help
Salman Al Zahrani Thanks for your informative comment. Regards, Humberto Santos
Dear @Humberto Santos ,
Did you build your model in 2D or 3D?
The magnitude of the back pressure at the outlet always affect the magnitude of pressure drop more than surface roughness.
What is your working fluid , does it liquid or gas?
What is the range of the temperature cross you have?
Where you are calculate the pressure drop, in tube side or shell side?
Regards.
- Badges
- Science topic