Hi,

I'm trying to automate the building of a complex compliant structural model that used static stabilization with the damping factor method. In my script I'd copied the damping factor magnitude from the existing model, which was using meter, kilogram, second units but I'm working in millimeter, kilogram, megapascal, etc units.

I was getting very problematic results, so I built a simple model comparable to the first step of the complex model. A shell-element cantilever beam is displaced by a contact interaction with a sphere.

With no damping the results are as expected and displacements and maximum stresses agree well with analytical beam bending formulas, suspect the the small (e.g. <2% for tip displacement) differences are because of the contact-loaded wide, thin beam I'm using.

with a damping factor 0.0002 (the original other model had 20! ) the result is a nonsensical localized deformation:

a damping factor of 2e-7 expands the area of sphere influence but still gives significant error.

If I use adaptive damping with the damping energy no more than 0.5% of the total strain energy, it works well. And I understand now to check ALLSD to ensure that the dissipation energy is not too high.

I'm pretty sure the larger complex model needs some damping to achieve convergence. The adaptive damping may be a reasonable choice, but I'd like to understand how to choose an appropriate value manually given some material properties and a chosen system of units.

Thanks,

Dan