3. Modeling A#
3.1. Characteristics of modeling#
We use a COQUE_AXIS model. The COQUE_AXIS models can be used both for thick plates (HENCKY - MINDLIN - REISSNER) and for thin plates (KIRCHOFF - LOVE) thanks to a penalization approach that makes it possible to neutralize or not the shear energy: this is the theory of HENCKY - MINDLIN - NAGHDI. In order to approach the LOVE - KIRCHHOFF solution numerically, it is necessary to take a sufficiently large shear coefficient (A_ CIS) to inhibit the transverse shear kinematics \({\mathrm{\gamma }}_{s}\). The larger this coefficient, the more the stiffness matrix is almost singular, the more the stiffness matrix is almost singular and therefore a source of numerical instabilities.
3.2. Characteristics of the mesh#
The mesh contains 100 elements of type SEG3.
3.3. Tested sizes and results#
We test the movement in the top left corner of the plate.
Identification |
Reference type |
Reference value |
Precision |
|
Point \(A\) - \(\mathrm{DX}\) |
“ANALYTIQUE” |
63.9488 |
|
|
Point \(B\) - \(\mathrm{DX}\) |
“ANALYTIQUE” |
32.000 |
|
|
Point \(C\) - \(\mathrm{DX}\) |
“ANALYTIQUE” |
0.05120 |
0.05 |
0.05 |
Point \(A\) - \(\mathit{DRZ}\) |
“ANALYTIQUE” |
0.06583 |
0.05 |
|
Point \(B\) - \(\mathit{DRZ}\) |
“ANALYTIQUE” |
41.133 |
|
|
Point \(B\) - \(\mathit{NYY}\) |
“ANALYTIQUE” |
2.0000 |
0.05 |
|
Point \(B1\) - \(\mathit{NYY}\) |
“ANALYTIQUE” |
3.84429 |
|
|
Point \(B1\) - \(\mathit{MXX}\) |
“ANALYTIQUE” |
-4.01497 10—2 |
|