Non linear loading curve failing during unloading

[Rue]

I am running an implicit analysis on a foam-filled structure. I am using a single curve for both the loading and unloading. The solver crashes during the unloading phase. Usually between 59% and 70% of the run time. The foam on its own loads and unloads following the prescribed paths. Could you please take a look at the keyword file and advise which inputs are incorrect, missing, or causing conflict? Any form of feedback on where I might be getting the physics wrong will be greatly appreciated.