A generalization of shear deformation theories for axisymmetric multilayered shells (Q1188008)

The problem of transverse shear (T.S.) deformations has been posed for long in the theory and the calculation of shell structures, first for rather thick shells and more recently for laminates, and led to break away partly from the Kirchhoff-Love (K-L) hypotheses. In this article, the author reconsiders the problem in the case of axisymmetric multilayered shells, and proposes a new generalized shear deformation theory. The kinematic assumptions classically consist in adding to the three- dimensional (3-D) K-L displacement, function of the meridional coordinate and the time, a supplementary displacement which takes account of the extrarotation of the normal. The total rotation is supposed here in the form of a sine function of the thickness coordinate orthogonal to the middle surface. The other displacements follow the standard assumptions for axisymmetric shells. All the theory is developed according to these assumptions and gives a variational principle. It makes conspicuous, after integrations with respect to the thickness, besides the classical surface stresses, not only a surface T.S. stress, but also, as would be expected, supplementary bending stresses functions of the assumed T.S. rotation. The surface equilibrium equations are also provided after integrations by parts. The author presents three approximations from an asymptotic expansion of the sine function, the K-L theory, the Reissner theory, and a third-order shear deformation theory, respectively. One of the main problems lies in the derivation of a correct surface constitutive equation, sufficiently accurate regarding the T.S. modulus. This was made classically by Reissner, then by Naghdi who found correction factors in assuming quadratic variations of the T.S. stress through the thickness for isotropic plates, and by Whitney for laminated plates. The choice of the T.S. rotation in a sine form is justified here in considering the 3-D plate equilibrium equations and results given by Cheng (1979) by an asymptotic expansion, which are restated without any approximation and which do not require the introduction of correction factors. Some applications are produced for particular axisymmetric shells which unfortunately suffer from comparisons with known exact solutions. Results corresponding to different approximations are actually compared to each other, and given as justifications of the new theory. Extension of the theory to arbitrary shapes is suggested and future works turned to edge effects will be studied.