A mathematical analysis of the simplest direct models for rods and shells (Q911830)

The purpose of this paper is to present a unified mathematical analysis of the so-called direct model of the shells and rods theory. The theory, based on Cosserat's approach, is revisited in a rather new way which pretends to reach a level of precision comparable to the three- dimensional analysis. In fact both shells and rods configurations can be represented locally by a disk seen as a stack of planes joined together by a bundle of filaments, or vice versa, each deforming possibly differently. A shell is just a two-dimensional set of disks, while a rod is a one-dimensional set of disks. After a detailed description of the kinematics of the local models based initially on one-field or two-fields directors for a shell or for a rod, respectively, and the corresponding images of disks, it is understood that the deformation between two local configurations corresponds to a particular deformation from a disk to another such that the stack of planes and filaments in the bundle are all deformed homogeneously. In fact that is true as a first approximation by which it is explained that all the deformations thus obtained are intrinsic, that is to say only planes and filaments are deformed homogeneously into planes and (straight) filaments. Classical restrictions on the small dimensions of the bodies, on the directors and on some regularity conditions are to be made. In another section, automorphisms of a disk and the concept of material symmetry at a point in a shell or in a rod, together with both reversible or non-reversible symmetries, are considered. Finally, in the last section the concept of global models is considered, where the global configuration is seen as a list of fields of deformations and stresses relative to the local model. From that viewpoint the case of material uniformity is explained as a particular case of a body materially isomorphic, but this is not too much developed in the present paper. To conclude, we are in view of a new representation of the direct method, which is rather discret on nonlinear theories, but which could give new ideas for the rest, even if it appears to some readers somewhat academic.