Divergent Flotran Solution

[zartan]

I am running a fairly simple 2D FLOTRAN analysis, and ANSYS keeps giving me an error when I run FLOTRAN saying that the "Coefficient Matrix has a negative diagonal." I've solved this same problem previously last week without any problems. But, since then I have not been able to get a solution. I actually haven't been able to solve any FLOTRAN problems because of this error for nearly a week. I was wondering if there was some minor thing I may be activating or deactivating that might cause this. Any help anyone could give me would be quite beneficial.

 

[zartan]

So, I get the correct solution when I use inches for units, and change the reference conditions accordingly. But, it won't solve with the same geometry and boundary conditions if I just use the equivalent meter lengths. It seems like I'm just missing some fine point here.

 

[mohr]

Hi zartan !!

Are you working well with your fluid DOF ?
Try with this
Main Menu>Solution>FLOTRAN Set Up>Relax/Stab/Cap>MIR Stabilization

and use in MIR Stabilization = 0.1 or 0.2 ( max 1.0)
and try the analysis again.
Also read "Relaxation factors" in Ansys help

This relaxation factor will help in conditioning the numerical operations avoiding "Coefficient Matrix has a negative diagonal."

Regards

 

[mohr]

Hi zartan !!

Are you working well with your fluid DOF ?
Try with this
Main Menu>Solution>FLOTRAN Set Up>Relax/Stab/Cap>MIR Stabilization

and use in MIR Stabilization = 0.1 or 0.2 ( max 1.0)
and try the analysis again.
Also read "Relaxation factors" in Ansys help

This relaxation factor will help in conditioning the numerical operations avoiding "Coefficient Matrix has a negative diagonal."

Regards
Mohr

 

[zartan]

Thanks, I'll look into those suggestions.

Z