3. Summary of results

The mesh is the one used for the « LISA » benchmark. The calculation was carried out until the moment \(t=2s\) . The regularized kinematic methods of EDF and the University of Liège give very similar results (difference less than 1% to \(t=2s\)). On the other hand, the lower bound provided by FZJ is more important than the previous upper bound.

Note: in this test case, we treat the “convergenceError” error by exception. Indeed, when the value of \(m\) approaches 1, the Norton-Hoff law tends towards the von Mises law. Knowing that the analysis of the limit load with the von Mises law is a singular problem that will cause non-convergence. Therefore, in case the calculation no longer converges with the growth of the moment \(t\) (and therefore when \(m\) tends to 1), the use of the exception makes it possible to recover the values of \(\mathrm{\lambda }\) calculated before the convergence difficulty.