[mazhar16823]

Hi everyone,


I am performing 3D steady state simulations using a circular domain for an airfoil and applying freestream BC at farfield with the help of STARCCM+ (Coupled Flow+Energy Solver and Gamma_ReTheta Model).



The coupled solver is used for this case, using grid sequencing initialization, boundary layer initialization, and the expert driver. A Courant number of 2000 is used, ramping from 0.1 over 200 iterations.



While running the simulation, a warning appears "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.
Command: Automation.Run"


And there is nothing mentioned in the user guide about this issue. Any idea why is this happening?

[bluebase]

Hi MA,


a floating point exception usually occurs when your have a division by zero, or (close-to) infinity.


Depending on your simulation, this might mean your simulation diverged (Check your residuals and flow quantities of interest).

In cases with somewhat complex user field functions, it could mean, you are using a field function which might be computed after an iteration, and has no values at the beginning at the iteration leading to a division by zero internally. This often needs one to take extra steps to deal with these cases.


By the way, the steve portal is always a good source of information regarding StarCCM. I'd consider posting in the respective subforum helpful as well.


Best,
Sebastian

[mazhar16823]

Originally Posted by bluebase

Hi MA,


a floating point exception usually occurs when your have a division by zero, or (close-to) infinity.


Depending on your simulation, this might mean your simulation diverged (Check your residuals and flow quantities of interest).

In cases with somewhat complex user field functions, it could mean, you are using a field function which might be computed after an iteration, and has no values at the beginning at the iteration leading to a division by zero internally. This often needs one to take extra steps to deal with these cases.


By the way, the steve portal is always a good source of information regarding StarCCM. I'd consider posting in the respective subforum helpful as well.


Best,
Sebastian

Hi Sebastien,
Thanks for your input. You're right for "a floating point exception usually occurs when your have a division by zero, or (close-to) infinity". Actually, I was using coupled flow+energy with Steady state once I moved to Implicit unsteady in an effort to avail adaptive time step - it works out and simulation has converged.