code_saturne User's Forum

Hi everyone,
I am comparing several meshes which utilise coupling CFD and experimental techniques against a fully resolved fine mesh model (19,000 elements using k epsilon model)

The coarse meshes are listed below

CMesh1-630 elements
CMesh2- 960 elements
CMesh3- 2400 elements
CMesh4- 3600 elements

What I have found out is that Cmesh1 produced the most accurate results compared to the fine mesh.

I am finding it difficult to explain this phenomenon. What I have learnt from my CFD lectures, is that a denser mesh provides more accurate results because it is more accurately interpolated between two points that are much closer together spatially.

Does anyone know why this could be the case, or point me in the direction of where to look, maybe mesh skewness plays a part.

Thank you in advance

Hello,

Mesh quality may play a part, but the main explanation may be error compensation.

Without the details of your computation, I can't say much, but your choice of wall law may be more or less adapted to a resolved mesh, and depending how features of your flow are captured by the k-epsilon model, some errors due to the model assumptions/approximations may partially compensate resolution/discretization error.

That is also why mesh sensitivity studies are important you may be "lucky" on a coarse mesh without the results proving anything...

Best regards,

Yvan

Thank you for the response.

I have conducted a mesh independent study where I started off with 42,000 elements and ended up with 19,000 elements, losing around 4% accuracy in my sensitivity study. To put things in perspective the pressure gradient at the end of the calculation for these two meshes was just over 1%.

The k epsilon advance options were set to scalable wall function that allowed y plus to go as low as around 11. My meshes y plus started at around 19 and ended at just under 40.

I have also added roughness to my model. There is little information provided, but I made sure my smallest element was larger then the roughness height given.

This last reason doesn't allow me to use the low Reynolds number models like SST k omega as the height of the first cell is much smaller then the roughness height. I believe my coarse meshes match well with the k epsilon model and my finer meshes would match well with a low Reynolds number model.

I have tried a mesh sensitivity study for SST model with roughness, but the discrepancy in the results fluctuates a lot.

I will try to investigate further to resolve this issue but thank you once again Yvan

  #6

asda3D

New Member

 

anonymous

Join Date: Apr 2018

Posts: 20

Rep Power: 3

Just an add on the aspect ratio is more than double for the finer meshes compared to the coarser meshes. This may be the reason for the coarser meshes possessing greater accuracy.