Bi-Conjugate Gradient Stabilized solver did not converge

wolfy_01 · September 3, 2020, 05:41

Hi CFD Lovers,

I am new to CFD world hence many of the technical i dont know. Can you please help me out with this issue.

I am doing internship trying to generate airfoils for BLI using PARSEC method. So my guide has simulated an architecture and now it seems that when I run his simulation it is saying

"WARNING: insufficient precision on multigrid level 15, nRows = 2
AMG coarsening halted.

Bi-Conjugate Gradient Stabilized solver did not converge !
A floating point exception has occurred: floating point exception [Invalid operation]. The specific cause cannot be identified. Please refer to the troubleshooting section of the User's Guide.
Context: star.coupledflow.CoupledImplicitSolver
Command: Automation.Run
error: Server Error".

Please see attached the images of the mesh to better understand the project.

Could you please me out with this!!

ping · September 4, 2020, 05:31

so you are saying that you have changed nothing and his previously working model will now no longer run - that would be very surprising unless maybe the version of star is different?

Kushal Puri · September 14, 2020, 07:00

Quote:

Originally Posted by wolfy_01

Hi CFD Lovers,

I am new to CFD world hence many of the technical i dont know. Can you please help me out with this issue.

I am doing internship trying to generate airfoils for BLI using PARSEC method. So my guide has simulated an architecture and now it seems that when I run his simulation it is saying

"WARNING: insufficient precision on multigrid level 15, nRows = 2
AMG coarsening halted.

Bi-Conjugate Gradient Stabilized solver did not converge !
A floating point exception has occurred: floating point exception [Invalid operation]. The specific cause cannot be identified. Please refer to the troubleshooting section of the User's Guide.
Context: star.coupledflow.CoupledImplicitSolver
Command: Automation.Run
error: Server Error".

Please see attached the images of the mesh to better understand the project.

Could you please me out with this!!

Just turn off the acceleration method in segregated flow -> pressure->Amg liner solver and try to re run the simulation.

wolfy_01 · September 26, 2020, 18:30

Thanks for the reply. I tried to reinitialize thats when everything went wrong. So without reinitializing the original file, I tried to run then it started to converge. I don't know what's the logic behind. . I dont know to tweak in StarCCM+ like to play with AMG solver and CFL number and to use other accelerators. Could you share your views and suggestions

ping · September 26, 2020, 23:31

it is very common to have to manually 'baby' along a cfd case when more advanced physics or complex flow features exist and what your experience probably indicates is that the previous converged run was started with more basic settings to get it going before changing mid-stream to the desired settings or even boundary conditions.
btw, if you know the steps taken to achieve this you can now automate it using a simulation operation.

September 4, 2020, 05:31		#2
ping Senior Member Ping Join Date: Mar 2009 Posts: 556 Rep Power: 20	so you are saying that you have changed nothing and his previously working model will now no longer run - that would be very surprising unless maybe the version of star is different? wolfy_01 likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
pisoFOAM (LES) - internal pipe flow - convergence	gu1	OpenFOAM Running, Solving & CFD	0	January 11, 2018 16:39
Periodic flow using Cyclic - comparison with Fluent	nusivares	OpenFOAM Running, Solving & CFD	30	December 12, 2017 05:35
Preconditioned Conjugate Gradient Solver!	Chidu	Main CFD Forum	1	January 21, 2004 11:43
Block Matrix Solver Using Conjugate Gradient	John	Main CFD Forum	2	July 31, 2001 12:08
Poisson eq and conjugate gradient method	Gary Dantinne	Main CFD Forum	2	March 19, 1999 00:24

September 26, 2020, 18:30		#4
wolfy_01 New Member swaroop Join Date: Sep 2020 Posts: 3 Rep Power: 5	Thanks for the reply. I tried to reinitialize thats when everything went wrong. So without reinitializing the original file, I tried to run then it started to converge. I don't know what's the logic behind. . I dont know to tweak in StarCCM+ like to play with AMG solver and CFL number and to use other accelerators. Could you share your views and suggestions

September 26, 2020, 23:31		#5
ping Senior Member Ping Join Date: Mar 2009 Posts: 556 Rep Power: 20	it is very common to have to manually 'baby' along a cfd case when more advanced physics or complex flow features exist and what your experience probably indicates is that the previous converged run was started with more basic settings to get it going before changing mid-stream to the desired settings or even boundary conditions. btw, if you know the steps taken to achieve this you can now automate it using a simulation operation.