Size of neighbour cells cause to solution error.
 

Hi!

I wonder why very good hexa mesh with two neighbour cell regions of very different sizes causes to terrible solution error. For example I have two neighbour layers of cells with heights 1 and 20. And there are non-physic results here.

Any suggestions?

Thanks.

Re: Size of neighbour cells lead to solution error
 

Look any CFD book for a Taylor series analysis of the error in case of different cell sizes. In my view a jump from 1 to 20 is too much, try reasonable clustering functions, hope that will help

Re: Size of neighbour cells lead to solution error
 

I'm sure it'll help. Thanks!