8. Summary of results

Very good results are observed in the presence of contact for both models.

The A, D and E models (DKT) present very satisfactory results compared to the analytical results (— 1% error with or without contact). When a plate under pressure is replaced by two plates, one of which sticks to the other, multiplying the stiffness by two results in a division by two of the maximum deflection as well as the stress at the center of the plates.

The B, C (COQUE_3D) models give a behavior equivalent to that encountered for the A, D and E models in DKT when replacing a plate under pressure with two plates in contact. We observe a division by two of the maximum deflection as well as of the stress at the center of the plates. In addition, the value of the arrow obtained is quite satisfactory (~ 7% error) compared to the analytical results. This difference can be explained by taking into account the transverse shear for COQUE_3D for a plate that is all in all relatively thick since \(l/e=1/10\).