|
[Sponsors] |
January 22, 2014, 02:08 |
Need help for the “FFD_SURFACE_POINTS”
|
#1 |
New Member
Yufei ZHANG
Join Date: Dec 2013
Posts: 5
Rep Power: 13 |
In the ONERA M6 optimization test case (V3.0.00 test case folder: TestCases\optimization_rans\steady_oneram6\), the grid file (mesh_ONERAM6_turb.su2) includes the FFD control parameters.
The “FFD_CORNER_POINTS” and “FFD_CONTROL_POINTS” is easy for understanding. But the “FFD_SURFACE_POINTS” is quite confusing. The point indexes are the indexes of the wing surface grid, but the coordinates are not on the wing surface (as shown in the figure). Is the FFD_SURFACE_POINTS necessary? Is there any tutorial for the FFD definition? When I use the FFD function with my own grid, the SU2_MDC couldn’t generate the deformed mesh, because the maximum difference of the original and deformed surface grids is very large, even when all of the geometry parameters are 0. I think the problem comes from the FFD_SURFACE_POINTS definition. Thank you for your help! |
|
January 23, 2014, 01:29 |
|
#2 |
New Member
Yufei ZHANG
Join Date: Dec 2013
Posts: 5
Rep Power: 13 |
It seems that there is a transformation relationship between the FFD_CONTROL_POINTS and the FFD_SURFACE_POINTS. If the FFD box is a unitary box (x, y, z) = ([0,1], [0,1], [0,1]), the output of SU2_MDC is correct when the coordinates of FFD_SURFACE_POINTS are equal to the surface grid points. If the FFD box is not a unitary box, the FFD_SURFACE_POINTS need to be transformed by a relation. But the relation is not clear.
|
|
January 28, 2014, 16:44 |
|
#3 |
Member
Trent Lukaczyk
Join Date: Feb 2011
Location: Stanford, CA
Posts: 75
Rep Power: 15 |
FFD_SURFACE_POINTS identify which surface grid points are inside the FFD box, and their (u,v,w) parametric location inside the box.
|
|
January 29, 2014, 20:51 |
|
#4 | |
New Member
Yufei ZHANG
Join Date: Dec 2013
Posts: 5
Rep Power: 13 |
Quote:
Thank you for your reply. How to define the (u,v,w) parametric location? Is there a transformation tool or code? Thank you! flyyufei |
||
January 30, 2014, 13:40 |
|
#5 | |
Member
Tommy Chen
Join Date: Mar 2011
Location: University of Michigan
Posts: 96
Rep Power: 15 |
Quote:
Are you from the Tsinghua University ? actually the FFD transformation algorithm could be found in a classicla paper, http://pages.cpsc.ucalgary.ca/~blob/...others/ffd.pdf and also http://www.cse.iitd.ernet.in/~pkalra/csl783/ch14ffd.pdf Cheers |
||
January 30, 2014, 13:55 |
|
#6 |
Member
Tommy Chen
Join Date: Mar 2011
Location: University of Michigan
Posts: 96
Rep Power: 15 |
Actually the parametric location is like sort of the (u,v) parameters in NURBS methods, to define the relatively location within the FFD control volume
|
|
February 5, 2014, 21:04 |
|
#7 | |
New Member
Yufei ZHANG
Join Date: Dec 2013
Posts: 5
Rep Power: 13 |
Hi Tommy,
Thank you very much for your help. I apologize for my late reply because of the Chinese Spring Festival. I’m from Tsinghua University in Beijing. How did you know? Quote:
|
||
February 12, 2014, 14:17 |
|
#8 |
Super Moderator
Thomas D. Economon
Join Date: Jan 2013
Location: Stanford, CA
Posts: 271
Rep Power: 14 |
Hi,
Currently, the code uses the 8 control points and the degree for each direction (i,j,k) to compute the locations of the FFD_CONTROL_POINTS (basically just the locations of each {i,j,k} of the structured box) and the FFD_SURFACE_POINTS automatically, i.e., no user intervention is required for FFD_CONTROL_POINTS or FFD_SURFACE_POINTS, only for FFD_CORNER_POINTS. Therefore, after creating your mesh and supplying the basic definition of the box (degree and corner points), run SU2_MDC with DV_KIND= NO_DEFORMATION in the configuration file, and the remaining information (FFD_CONTROL_POINTS and FFD_SURFACE_POINTS) will be populated automatically for you. All the best, Tom |
|
February 20, 2014, 07:31 |
|
#9 | |
New Member
Yufei ZHANG
Join Date: Dec 2013
Posts: 5
Rep Power: 13 |
Thank you! It works!
Quote:
|
||
|
|