problems with winglet shape optimization, cannot impose geometrical constrains

lpustin · January 31, 2018, 06:14

Hello all,
I'm working with SU2 for my master thesis.
I have already performed an Euler analysis of a wing with a winglet.
Now I would like to perform an optimization also of the winglet shape but I have problems with the input file, and precisely, on the geometrical constraint section.
Below are shown some lines of the original Onera M6 optimization test-case (inside the SU2 test-cases) configuration file to underline my problem.

% Orientation of airfoil sections (X_AXIS, Y_AXIS, Z_AXIS)
GEO_AXIS_STATIONS= Y_AXIS
%
% Coordinate of the sections
GEO_LOCATION_STATIONS= (0.0, 0.2, 0.4, 0.6, 0.8)
%
% Location (coordinate) of the airfoil sections (MinValue, MaxValue)
GEO_WING_BOUNDS= (0.0806, 1.1284)

% Optimization constraint functions with scaling factors, separated by semicolons
% ex= (Objective = Value ) * Scale, use '>','<','='
OPT_CONSTRAINT= (LIFT > 0.2864) * 10.0; (MAX_THICKNESS_SEC1 > 0.0570) * 0.1; (MAX_THICKNESS_SEC2 > 0.0513) * 0.1; (MAX_THICKNESS_SEC3 > 0.0457) * 0.1; (MAX_THICKNESS_SEC4 > 0.0399) * 0.1; (MAX_THICKNESS_SEC5 > 0.0343) * 0.1

Then, basically, I can provide the location of the airfoil section inside the wing bounds and have SU2 performing the optimization with the max thickness constrains.

My problem is that is possible to specify only X, Y or Z axis as orientation for the airfoil and, obviously, a winglet is not planar and thus, cannot comply with having the airfoil section perpendicular to the X,Y or Z orientation.
I've tried to perform optimization without constrains on the winglet thickness but I've obtained self-intersecting surfaces.

Is there any way to give reasonable constraints on a winglet ?
Hope somebody could help me,
Thanks

fpalacios · February 24, 2018, 12:49

What you are proposing makes sense but it requires some implementation.
First of all I would check the new tutorials for the SU2 v6.0 (the ONERA optimization is applicable).
Then I'll modify some lines of the code in particular how the wing slicing is done. Search for
config->GetGeo_Description() == WING and you will find how the normal plane is defined using

Plane_Normal[0] = ___
Plane_Normal[1] = ___
Plane_Normal[2] = ___

By the way, if you are planning to develop SU2... take a look at www.su2devsociety.org

Best,
Francisco

lpustin · February 26, 2018, 07:02

Thanks you very much Francisco.
We implemented changes to the code to allow for nonplanar wing shape optimization maintaining a C1 continuity between adjacent FFD boxes (and without freezing in space the relative Control Points).

We would like to share this effort with the community of SU2, thus, I will follow your suggestion regarding the SU2dev society.

January 31, 2018, 06:14	problems with winglet shape optimization, cannot impose geometrical constrains	#1
lpustin New Member Luca Join Date: Nov 2017 Location: Rome Posts: 5 Rep Power: 8	Hello all, I'm working with SU2 for my master thesis. I have already performed an Euler analysis of a wing with a winglet. Now I would like to perform an optimization also of the winglet shape but I have problems with the input file, and precisely, on the geometrical constraint section. Below are shown some lines of the original Onera M6 optimization test-case (inside the SU2 test-cases) configuration file to underline my problem. % Orientation of airfoil sections (X_AXIS, Y_AXIS, Z_AXIS) GEO_AXIS_STATIONS= Y_AXIS % % Coordinate of the sections GEO_LOCATION_STATIONS= (0.0, 0.2, 0.4, 0.6, 0.8) % % Location (coordinate) of the airfoil sections (MinValue, MaxValue) GEO_WING_BOUNDS= (0.0806, 1.1284) % Optimization constraint functions with scaling factors, separated by semicolons % ex= (Objective = Value ) * Scale, use '>','<','=' OPT_CONSTRAINT= (LIFT > 0.2864) * 10.0; (MAX_THICKNESS_SEC1 > 0.0570) * 0.1; (MAX_THICKNESS_SEC2 > 0.0513) * 0.1; (MAX_THICKNESS_SEC3 > 0.0457) * 0.1; (MAX_THICKNESS_SEC4 > 0.0399) * 0.1; (MAX_THICKNESS_SEC5 > 0.0343) * 0.1 Then, basically, I can provide the location of the airfoil section inside the wing bounds and have SU2 performing the optimization with the max thickness constrains. My problem is that is possible to specify only X, Y or Z axis as orientation for the airfoil and, obviously, a winglet is not planar and thus, cannot comply with having the airfoil section perpendicular to the X,Y or Z orientation. I've tried to perform optimization without constrains on the winglet thickness but I've obtained self-intersecting surfaces. Is there any way to give reasonable constraints on a winglet ? Hope somebody could help me, Thanks

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[ICEM] Problems with coedge curves and surfaces	tommymoose	ANSYS Meshing & Geometry	6	December 1, 2020 11:12
[ANSYS Meshing] Wing with winglet mesh problems	MAH_cape	ANSYS Meshing & Geometry	1	May 31, 2016 04:06
Error during shape optimization	Ullixes91	SU2	3	February 23, 2016 02:45
Shape Optimization Error, Incorrect values produces by SU2_GDC	Dhig	SU2 Shape Design	1	March 20, 2014 05:58
Shape optimization	chuls	Main CFD Forum	1	October 6, 2009 04:00

February 24, 2018, 12:49		#2
fpalacios Super Moderator Francisco Palacios Join Date: Jan 2013 Location: Long Beach, CA Posts: 404 Rep Power: 15	What you are proposing makes sense but it requires some implementation. First of all I would check the new tutorials for the SU2 v6.0 (the ONERA optimization is applicable). Then I'll modify some lines of the code in particular how the wing slicing is done. Search for config->GetGeo_Description() == WING and you will find how the normal plane is defined using Plane_Normal[0] = ___ Plane_Normal[1] = ___ Plane_Normal[2] = ___ By the way, if you are planning to develop SU2... take a look at www.su2devsociety.org Best, Francisco

February 26, 2018, 07:02		#3
lpustin New Member Luca Join Date: Nov 2017 Location: Rome Posts: 5 Rep Power: 8	Thanks you very much Francisco. We implemented changes to the code to allow for nonplanar wing shape optimization maintaining a C1 continuity between adjacent FFD boxes (and without freezing in space the relative Control Points). We would like to share this effort with the community of SU2, thus, I will follow your suggestion regarding the SU2dev society.