[msv_kk]

Hello,

My name is Kiran Kumar,Meda, I am doing my Masters at Duisburg -Essen univeristy. Currently I am working on my Thesis which is related to CFD.

I am doing a Grid independent study, I tried to generate three different meshes using Different absolute minimum surface size, form which I would like to select one for my furthure analysis (Validation with Test results).

I am using Ideal Gas , Segregated flow solver, realizable K-Epsilpon model with Two-layer all y+ treatment.

Here is the info about the CFD simulation results,

Min sur size Number of cells Temp o/l (°c )
0,6 5262106 79,29
0,4 8345705 84,32
0,3 10565895 83,45
The test resule is 81,6 (°c).I would like to continue with the mesh having 0,4 mm as the min surface size.


Now, I had a discussion with my friend, who says ' as the mesh size decreses the resistance to flow decrease, which results in less friction ,less turbulence intensity. So the temperature at the oulet for mesh (0.3) should be higher than mesh (0.4)'.

Is this argument correct ? he also says as we go on reducing the min surface size, the resitance to flow decreases which results in high temperature at the outlet.


Thanking you

Best Regards,
Kiran Kumar,Meda

[sail]

Originally Posted by msv_kk

Hello,

My name is Kiran ....

an increase in 10% (cell count between 0.3 and 0.4) is not considered sufficient to analyse the grid dependency of a simulation. while the sacred books recommend a 2x for every step, it is not always possible and an increment of 50% is usually tolerated.

i don't really understand your friend point. about winch velocity and resistance is he talking about? ideally there shoudent' be any differences in the flow.

[abdul099]

Additionally to sail's post, I suggest to modify the base size instead of the min surface size. The min surface size affects the mesh only on faces with strong curvature or with close proximity to other faces. All other faces will be meshed with the target surface size, so it might not be meaningful to reduce only the min surface size.

And it's nearly impossible to give any statement whether it makes sense with higher / lower / whatever temperatures when modifying any mesh sizes. I don't even know if the gas should cool down in the domain or should heat up. So it's impossible to give any statement!