|
[Sponsors] |
November 24, 2018, 16:40 |
Unstructured mesh vs Cartesian cutcell
|
#1 |
Senior Member
Selig
Join Date: Jul 2016
Posts: 213
Rep Power: 10 |
I was curious what perspectives,opinions, and experiences people have had with unstructured solvers and cartesian cut cell based solvers.
Will cartesian cut cell based solvers rival unstructured solvers? A definitive advantage of cartesian cut cell methods is you can use high-order methods such as ADER and WENO. |
|
November 24, 2018, 22:30 |
|
#2 |
Super Moderator
|
High order ADER and WENO have been done on unstructured also. Maybe some algorithms are more efficient on Cartesian and also better utilization of computer resources due to fixed Cartesian stencils.
But boundary conditions, especially for RANS will be a challenge for cut cell methods. May be you can improve these things but it will require you to do more algorithm development and implementation issues. Each method has its own advantages and disadvantages. I think it is not possible to say which one is better overall, because to make fair comparison, you will have to invest a huge amount of time to develop and implement both with great care. We cannot judge such things based on theoretical or academic grounds, but rigorously test on real world problems. At this stage so many other issues will crop up that some theoretical advantage of one method may become irrelevant. So much developments have been done in each class of methods, that I dont think that any one method will be able to rival others in future. I feel there is a "conservation of difficulty"; you switch from one method to another, some things become easier, others become harder. |
|
November 25, 2018, 03:15 |
|
#3 |
Senior Member
Filippo Maria Denaro
Join Date: Jul 2010
Posts: 6,783
Rep Power: 71 |
I think that cut-cell method is now supersided by the immersed boundary method. Honestly, I think that both are inaccurate close to the boundaries to be suitable for simulations like DNS/LES. Maybe, wall -modelled BCs can somehow take into account for the issue. RANS could be more suitable.
I think that unstructured grids allows for superior methods. |
|
November 25, 2018, 12:52 |
|
#4 |
Senior Member
|
Let me add that the question is not exactly well posed.
Unstructured grid is a grid which is not structured. This says nothing on the method used to construct it nor the type of cells used, presence or not of boundary layer, etc. Cartesian cut cell (ccc), in contrast, is formally a meshing technique, which, typically, actually ends up producing an unstructured grid; that is, the underlying cartesian grid structure is not actually retained in the solver. Also, if by ccc you refer to cutting the cartesian grid with the geometry without further enhancements, this is not how this is typically implemented, as it is just the first of several steps. If we compare this first step with a well designed grid with a boundary layer, there is no story of course. However, this first ccc step alone is typically not working and requires several immersed boundary like tricks to avoid crashes. |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
how to set periodic boundary conditions | Ganesh | FLUENT | 15 | November 18, 2020 06:09 |
unstructured mesh generation with strictly constrained boundary in simple space | goujl | Mesh Generation & Pre-Processing | 0 | January 29, 2015 10:44 |
Moving mesh | Niklas Wikstrom (Wikstrom) | OpenFOAM Running, Solving & CFD | 122 | June 15, 2014 06:20 |
Mesh motion with Translation & Rotation | Doginal | CFX | 2 | January 12, 2014 06:21 |
unstructured cartesian mesh generation | hadian | Main CFD Forum | 3 | October 14, 2009 17:54 |