New obj func in discrete adjoint solver

Zhen ZHANG · August 1, 2021, 22:27

Hi all,

I use a new objective function in the discrete adjoint solver. But the derivative of this new objective function is about 1e-11 and much smaller than of other existed objective functions.

Should I do anything to make this new objective function available to the discrete solver?

Thanks!

Zhen ZHANG · August 1, 2021, 23:17

I got it!

In CFVMFlowSolverBase<V,R>::EvaluateCommonObjFunc(con st CConfi
g& config) const, the new obj func should be added. Otherwise, the obj func in the discrete solver will be zero, and the derivative is not correct.

August 1, 2021, 22:27	New obj func in discrete adjoint solver	#1
Zhen ZHANG Member Zhen ZHANG Join Date: Jun 2018 Location: Beijing, China Posts: 56 Rep Power: 8	Hi all, I use a new objective function in the discrete adjoint solver. But the derivative of this new objective function is about 1e-11 and much smaller than of other existed objective functions. Should I do anything to make this new objective function available to the discrete solver? Thanks!

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
foam-extend-4.1 release	hjasak	OpenFOAM Announcements from Other Sources	19	July 16, 2021 05:02
the gradients calculated through discrete adjoint are near zero	Haowang	SU2	9	October 17, 2019 22:53
Adjoint Solver: Bounded by Surfaces constraint not converging	J.J	FLUENT	0	March 13, 2019 19:56
Star cd es-ice solver error	ernarasimman	STAR-CD	2	September 12, 2014 00:01
A question about discrete adjoint method	Thomas	Main CFD Forum	3	August 9, 2006 12:37

August 1, 2021, 23:17		#2
Zhen ZHANG Member Zhen ZHANG Join Date: Jun 2018 Location: Beijing, China Posts: 56 Rep Power: 8	I got it! In CFVMFlowSolverBase<V,R>::EvaluateCommonObjFunc(con st CConfi g& config) const, the new obj func should be added. Otherwise, the obj func in the discrete solver will be zero, and the derivative is not correct.