Troubleshooting NaN Values in MoorDyn: Incorrect Line Connection or Free Point Issue?

[wufangyun92]

Hi, I'm encountering the issue shown in the attached image. During the simulation, MoorDyn reports NaN values at specific time steps, and the simulation ultimately fails with a "NaN state detected" error.

I'm not sure whether the issue is caused by an incorrect mooring line setup (e.g., line type, attachment points) or if it's due to a Free type point that isn't properly connected to the system. Could this be why NaN values are appearing in the fairlead forces and tensions?

Any guidance on how to troubleshoot this would be greatly appreciated!

[RyanDavies19]

Hi @wufangyun92,

On first glance it seems like your line and point definitions are fine. MoorDyn throwing Nan's is a sign of numerical instability. What does your solver options list look like?

Common causes of instability in the initialization stage is too large of a dtM (moordyn timestep) value or also too high of a CdScaleIC (drag scaling) value.

[wufangyun92]

But dtM and CdScaleIC are both preset values. If so, do they need to be smaller?

[RyanDavies19]

Yes, I would try first reducing CdScaleIC as that will preserve the runtime efficiency (though it will take longer to find steady state during initialization). If that doesn't work, I would try reducing your dtM to 0.0001