Learning Forum

[Dominik Buksa]

Hello. I have a question about the Peclet number. My geometry is very irregular and I would like to know if there is advection anywhere in the flow. How can I calculate the Peclet number? Or can ansys count it? Should the characteristic dimension of this number be the size of the cell?

[Rob]

Have a look at the full Peclet Number definition re characteristic dimension. Fluent has functions to calculate pretty much anything of interest, but you'd need to add the length scale into that yourself as these are determined at the cell level, and the cells don't have any information other than for that cell. Ie there is no domain diameter/length etc value in the cell. 

[Dominik Buksa]

I understand, but each cell has a different characteristic dimension and a different local velocity value. Can I read, for example, the average or something in the fluent so that I can get the result?
I understand the approach to each cell but with a million elements it is a big problem.

[Dominik Buksa]

Can ansys tell me e.g. the smallest cells and their dimensions? So I could manually count the Perclet number at critical places.

[Rob]

How would you calculate the value in an experiment? That's usually the better way to approach nondimensional numbers with a length scale in simulation: cell size is very variable in most models. 

[Dominik Buksa]

Only for me, in each place of the flow, the number will be different and I don't know how to check the convective-advective condition? I put my geometry in the photo (the inlet is only 0.2m and the geometry cross-section is from 2m to 3.5m...). Do you have any idea how to solve it?

[Rob]

What do you need the number for? Ie experimental comparison. Dimensionless numbers tend to be used to compare designs so you "just" need to pick a length scale that makes sense for your needs. 

[Dominik Buksa]

I wanted to calculate the Pereclet number because I read on CFD forums that it must be less than 2 for the solution to be stable.
Additionally, I wanted to know what part is advection and what part is diffusion.

[Rob]

I'd not seen that one. I tend to go with, are the monitor points changing, how are the residuals and do the fluxes balance.