9. Synthesis

The results obtained with modeling A correspond well to the analytical solution proposed based on the theory of beams. This therefore validates the choice of the beam model as a reference.

In the B to E models (DKT and Q4GG), the stress values in the cable at the output of CALC_PRECONT are indeed those expected. On the other hand, the arrow values obtained are almost identical to the analytical solution. We note that the percentage of error is less important with modeling DKT for this case study, but especially that the error on the arrow is much lower than with modeling \(\mathrm{3D}\).

The F modeling makes it possible to validate that the prestress calculation is correct when the concrete is modeled by law GLRC_DAMAGE.

This validates the use of macro command CALC_PRECONT for shells.