Adjoint solution diverging with "Frozen viscosity" option as "No"

diwakaranant · October 7, 2013, 14:53

Hi

I was running adjoint simulation for RANS case for Onera M6 wing on hybrid coarse mesh.

The adjoint simulation is converging with "Frozen viscosity" option as "Yes"
but it is diverging when the option is kept as "No".

Can anyone say why is this happening ?

I am attaching the convergence plot and the .cfg file

Thanks
Anant

economon · October 11, 2013, 15:49

Hi Anant,

Setting the frozen viscosity to NO will enable the adjoint turbulence model (S-A) that is implemented in the code. Including the sensitivity of the viscosity by coupling the adjoint turbulence model can often slow convergence (not to mention that the stability requirements will likely be different too). You might try modifying the numerical methods and CFL number in this scenario to see if you can regain convergence.

Hope this helps,
Tom

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Problems with CEL (guess it's simple to solve)	Felggv	CFX	22	March 26, 2019 16:42
Domain Imbalance	HMR	CFX	5	October 10, 2016 05:57
Error finding variable "THERMX"	sunilpatil	CFX	8	April 26, 2013 07:00
forced to sticking of soot particle	kmgraju	CFX	0	November 27, 2012 09:08

October 11, 2013, 15:49		#2
economon Super Moderator Thomas D. Economon Join Date: Jan 2013 Location: Stanford, CA Posts: 271 Rep Power: 14	Hi Anant, Setting the frozen viscosity to NO will enable the adjoint turbulence model (S-A) that is implemented in the code. Including the sensitivity of the viscosity by coupling the adjoint turbulence model can often slow convergence (not to mention that the stability requirements will likely be different too). You might try modifying the numerical methods and CFL number in this scenario to see if you can regain convergence. Hope this helps, Tom