11. Summary of results

These tests, which aim to validate the law of behavior of concrete BETON_REGLE_PR, also serve to show a certain number of its weaknesses. The various models make it possible to test:

A and E: traction/compression behavior

Bet F: The behavior in cyclic flexure

This G: the behavior combining membrane and flexure phenomena

D:The behavior for shear and in-plane distortion

H: a thermal load.

In most of these situations, the efforts and moments predicted by the BETON_REGLE_PR model are consistent with the ENDO_ISOT_BETON model for the load. In case of discharge, model BETON_REGLE_PR, which is elastic, has significant errors compared to model ENDO_ISOT_BETON. Moreover, since the Poisson effect is not modelled by model BETON_REGLE_PR, significant differences in displacements perpendicular to the loading direction are measured.

Compared to the ssns106 test case, the flexure and shear coupling case in the plane is not modelled here (D and M models of the ssns106 test case). Indeed, there are difficulties in convergence with law BETON_REGLE_PR for these cases.