CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > Software User Forums > SU2

Geometric Constraints and Gradients for Constrained Wing Optimization

Register Blogs Community New Posts Updated Threads Search

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   January 24, 2014, 15:38
Default Geometric Constraints and Gradients for Constrained Wing Optimization
  #1
Member
 
Tommy Chen
Join Date: Mar 2011
Location: University of Michigan
Posts: 96
Rep Power: 15
Tommy Chen is on a distinguished road
Hi Stanford Guys

in the Tutorial 8
http://adl-public.stanford.edu/docs/...f+a+Fixed+Wing

There are thickness constraints along 5 sections of the wing, which are
MAX_THICKNESS_SEC1-5
However , I do not understand how do the solver determine the locations along the span direction of the wing for these 5 Thickness constraints.

Besides, I am reading the python scripts , I found that in the SU2/eval/gradients.py
the part of codes for
Mesh updating , mesh decomposition and deformation, and geometry solution
has been commented. like below:

##################################
# in SU2/eval/gradients.py

# ----------------------------------------------------
# Update Mesh (check with Trent)
# ----------------------------------------------------

# does decomposition and deformation
# info = update_mesh(config,state)


# ----------------------------------------------------
# Geometry Solution
# ----------------------------------------------------

##################################

Then how does SU2 calculate the geometric gradients, and how to achieve the geometry constraints like thickness??
Tommy Chen is offline   Reply With Quote

Old   January 28, 2014, 16:39
Default
  #2
Member
 
Trent Lukaczyk
Join Date: Feb 2011
Location: Stanford, CA
Posts: 75
Rep Power: 15
rktchip is on a distinguished road
haha nice. yup that line is supposed to be commented. i'm not sure how the code determines the span locations (maybe it's evenly spacing them along the monitors markers in the y direction?). It calculates surface sensitivities to geometry by finite difference inside SU2_GDC, using the same routines as SU2_MDC
rktchip is offline   Reply With Quote

Old   January 28, 2014, 22:06
Default
  #3
Member
 
Tommy Chen
Join Date: Mar 2011
Location: University of Michigan
Posts: 96
Rep Power: 15
Tommy Chen is on a distinguished road
Quote:
Originally Posted by rktchip View Post
haha nice. yup that line is supposed to be commented. i'm not sure how the code determines the span locations (maybe it's evenly spacing them along the monitors markers in the y direction?). It calculates surface sensitivities to geometry by finite difference inside SU2_GDC, using the same routines as SU2_MDC
Hi Trent

Thank you for your reply.

I have another question, I found out that the mesh deformation code has to be coupled with the design variables, suppose that I already have a deformed surface mesh, and I want to use SU2_MDC to get the deformed volume SU2 mesh with either spring or FEA method. Is that possible ? If it is, how can I do that?

Thanks
Tommy Chen is offline   Reply With Quote

Old   January 29, 2014, 03:09
Default
  #4
Member
 
Trent Lukaczyk
Join Date: Feb 2011
Location: Stanford, CA
Posts: 75
Rep Power: 15
rktchip is on a distinguished road
if i understand your question right, SU2_MDC deforms both the surface and the volume mesh. The *.su2 file that comes out is the deformed volume mesh (including the markers that identify the surface)
rktchip is offline   Reply With Quote

Old   January 29, 2014, 09:20
Default
  #5
Member
 
Tommy Chen
Join Date: Mar 2011
Location: University of Michigan
Posts: 96
Rep Power: 15
Tommy Chen is on a distinguished road
Quote:
Originally Posted by rktchip View Post
if i understand your question right, SU2_MDC deforms both the surface and the volume mesh. The *.su2 file that comes out is the deformed volume mesh (including the markers that identify the surface)
Right, that also means the users can only chose to parameterize the surface mesh by the existing methods within SU2. So , in SU2_MDC.cpp

################################################## ###
if (rank == MASTER_NODE) cout << endl << "------------------------- Surface grid deformation ----------------------" << endl;

/*--- Definition and initialization of the surface deformation class ---*/

surface_movement = new CSurfaceMovement();
surface_movement->CopyBoundary(geometry[ZONE_0], config[ZONE_0]);

/*--- Surface grid deformation ---*/

if (rank == MASTER_NODE) cout << "Performing the deformation of the surface grid." << endl;
surface_movement->SetSurface_Deformation(geometry[ZONE_0], config[ZONE_0]);

#ifndef NO_MPI
/*--- MPI syncronization point ---*/
#ifdef WINDOWS
MPI_Barrier(MPI_COMM_WORLD);
#else
MPI::COMM_WORLD.Barrier();
#endif
#endif
################################################## ##################

If I modify this line :
surface_movement->SetSurface_Deformation(geometry[ZONE_0], config[ZONE_0]);

can I make the SU2_MDC read some existing deformed surface mesh from file instead of deforming it within SU2_MDC?

Thanks Trent
Tommy Chen is offline   Reply With Quote

Old   January 29, 2014, 12:00
Default
  #6
Member
 
Tommy Chen
Join Date: Mar 2011
Location: University of Michigan
Posts: 96
Rep Power: 15
Tommy Chen is on a distinguished road
Quote:
Originally Posted by rktchip View Post
if i understand your question right, SU2_MDC deforms both the surface and the volume mesh. The *.su2 file that comes out is the deformed volume mesh (including the markers that identify the surface)
Never mind

I found that this could be achieved by defining the DV_KIND=surface_file and passing the MOTION_FILENAME of the deformed surface file to it

cheers
Tommy Chen is offline   Reply With Quote

Reply


Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are On
Refbacks are On



All times are GMT -4. The time now is 14:08.