3. Modeling A#
3.1. Characteristics of modeling#
3.2. Characteristics of the mesh#
Number of knots: 224
Number of meshes and type: 192 QUAD4
3.3. Tested values#
Identification |
Numerical reference [bib1] |
Experimental reference [bib2] |
Aster |
% differences |
||
Move \(w\) to point \(F\) |
|
|
|
|||
Move \(w\) to point \(C\) |
\(–0.7{10}^{\mathrm{-}3}m\) |
|
|
|||
Move \(w\) to point \(D\) |
|
|
|
|||
Constraint \(\mathit{SIXX}\) at point \(F\) |
|
|
|
|||
Constraint \(\mathit{SIYY}\) at point \(F\) |
|
|
|
3.4. Value of the normal displacement W along CD#

It can be noted that beyond the differences observed on the tested points \(C\), \(F\), \(D\), the normal displacement calculated along \(\mathit{CD}\) is close to the theoretically « deep shells » solution adopted by Batoz [bib1]. The errors relating to points \(F\) and \(D\) can be attributed to the low value of the displacement (of the order of \({10}^{\mathrm{-}5}m\)).
3.5. Stress value along CD#


It can be noted that the constraints calculated along \(\mathrm{CD}\) are generally in agreement with the theoretically « deep shells » solution adopted by Batoz [bib1].
3.6. notes#
Values for the CISA_L and CISA_T coefficients are not available. As the structure is thin (\(h\mathrm{/}R\mathrm{=}0.045\)), it is assumed that the effects of transverse shear are negligible, so we imposed CISA_L = CISA_T = \({10}^{10}\).
The normal displacement \(w\) (figure of [§4.2]) is expressed in the local cylindrical coordinate system \((R,\theta ,z)\), it is the displacement normal to the shell element. The displacement \(w\) tested in [§4.1] is for its part expressed in the global coordinate system (displacement following \(z\)).