scientific article; zbMATH DE number 4005709
[For the entire collection see Zbl 0592.00032.] In this article, the author rightly recalls that the introduction of fully nonlinear terms in change of curvature tensors in shell theories leads to virtually useless applications, but contrarily that engineering approximations may lead to inconsistent equations. The need of accurate numerical analyses gave rise of many different theories. In a pure Lagrangian theory it is proposed a so-called large-small approximation, where rotations about tangents of the middle surface may be large, while rotations about the normal remain small. The shell is supposed to be thin, isotropic and elastic and restricted to Kirchhoff-Love assumptions, and in the given theoretical developments some other terms may be neglected providing worse approximations, up to the classical moderate rotations theory. A first comparison of the different theories given in the literature is made in the case of cylindrical bending in strain measures, which show the advantage of the large-small rotations theory Then the comparison is made on one-dimensional and two-dimensional problems. In the first ones, namely a non-symmetrical, loaded, infinite deep, circular panel, or tori undergoing a ring load with varying principal radius of curvature, the numerical results giving the load- deflection curves of the different theories are compared to the exact solution. In all cases the large-small approximation results are close to the exact solution, while deviations often large are obtained in the others. Two two-dimensional problems are also presented, a diametrical, point-loaded, cylindrical tube and a point-loaded cylindrical panel, and solved by the F.E.M. method through a CDC Cyber 205 computer using a modern vectorized shell element. The results of the proposed theory reveal to be in full agreement with those of the most refined available one. In conclusion, the advantages of the large-small approximation version seem to be confirmed by the given examples up to large rotations. An interesting question incidentally arises in the numerical analysis: why the presence of second derivatives of the tangential displacement makes the iteration convergence easier? As a general consideration, would mixed formulations theories with finite rotations as independent unknowns bring any solution to this kind of problems? [see \textit{L. Recke} and \textit{W. Wunderlich}, ibid. 239-258 (1986; Zbl 0595.73084) or the reviewer, ibid. 317-332 (1986; Zbl 0595.73039)].
- On the derivation and comparative analysis of large rotation shell theories
- scientific article; zbMATH DE number 3878480
- A current trend in shell theory: Constrained geometrically nonlinear Kirchhoff-Love type theories based on polar decomposition of strains and rotations
- scientific article; zbMATH DE number 3907925
- scientific article; zbMATH DE number 3858580
- scientific article; zbMATH DE number 3858580 (Why is no real title available?)
- Vecteur de rotation incrémental pour les coques non linéaires en grandes rotations
- On the derivation and comparative analysis of large rotation shell theories
- scientific article; zbMATH DE number 4031136 (Why is no real title available?)
- Shell theory versus degeneration—a comparison in large rotation finite element analysis
- Allowing for Rotations about the Normal in Nonlinear Theories of Shells
- OWL: a code for the two-center shell model with spherical Woods-Saxon potentials
This page was built for publication:
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3757450)