Inverse design Cp file

yugang881 · December 7, 2017, 23:50

Hi dear friends:
I want to use SU2 to do inverse design. I find in the test case the inverse design Cp file also including the P and Mach number. I delete the P and Mach number and it runs to a very thin airfoil.
I am wondering how could i do the inverse design if i only have the target Cp? Could anyone show me the way? Thanks in advance!

Zen · December 8, 2017, 08:36

Hi James,

Indeed you don't need the pressure and the Mach number to run the inverse design optimization. If you want to have a look how the objective function for the inverse design is computed you can have a look at the code in output_structure.cpp under:

Code:

COutput::SetCp_InverseDesign(...)

or, alternatively, if you are interested in its variation (the one used in the adjoint code) you can have a look in solver_adjoint_mean.cpp uner:

Code:

CAdjEulerSolver::SetForceProj_Vector(...)

I think there are some small bugs in COutput::SetCp_InverseDesign(...), especially if you run it in parallel or if you specified your objective function on one boundary (e.g. an airfoil) but you have other boundaries of the same kind (e.g. solid walls).

Hope this helps,

Z

yugang881 · December 9, 2017, 05:03

Hi Zen:
Thanks for you kindly reply. I will check the code soon.

December 7, 2017, 23:50	Inverse design Cp file	#1
yugang881 New Member JAMES Join Date: Dec 2016 Posts: 21 Rep Power: 10	Hi dear friends: I want to use SU2 to do inverse design. I find in the test case the inverse design Cp file also including the P and Mach number. I delete the P and Mach number and it runs to a very thin airfoil. I am wondering how could i do the inverse design if i only have the target Cp? Could anyone show me the way? Thanks in advance!

December 8, 2017, 08:36		#2
Zen Member Zeno Join Date: Sep 2013 Location: Delft, The Netherlands Posts: 63 Rep Power: 13	Hi James, Indeed you don't need the pressure and the Mach number to run the inverse design optimization. If you want to have a look how the objective function for the inverse design is computed you can have a look at the code in output_structure.cpp under: Code: COutput::SetCp_InverseDesign(...) or, alternatively, if you are interested in its variation (the one used in the adjoint code) you can have a look in solver_adjoint_mean.cpp uner: Code: CAdjEulerSolver::SetForceProj_Vector(...) I think there are some small bugs in COutput::SetCp_InverseDesign(...), especially if you run it in parallel or if you specified your objective function on one boundary (e.g. an airfoil) but you have other boundaries of the same kind (e.g. solid walls). Hope this helps, Z

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[swak4Foam] Problem installing swak_2.x for OpenFoam-2.4.0	towanda	OpenFOAM Community Contributions	6	September 5, 2015 22:03
centOS 5.6 : paraFoam not working	yossi	OpenFOAM Installation	2	October 9, 2013 02:41
"parabolicVelocity" in OpenFoam 2.1.0 ?	sawyer86	OpenFOAM Running, Solving & CFD	21	February 7, 2012 12:44
Version 15 on Mac OS X	gschaider	OpenFOAM Installation	113	December 2, 2009 11:23
ParaView Compilation	jakaranda	OpenFOAM Installation	3	October 27, 2008 12:46

December 9, 2017, 05:03		#3
yugang881 New Member JAMES Join Date: Dec 2016 Posts: 21 Rep Power: 10	Hi Zen: Thanks for you kindly reply. I will check the code soon.