End effects for anti-plane shear deformations of sandwich structures

The purpose of this research is to further investigate the effects of material inhomogeneity and the combined effects of material inhomogeneity and anisotropy on the decay of Saint-Venant end effects. Saint-Venant decay rates for self-equilibrated edge loads in symmetric sandwich structures are examined in the context of anti-plane shear for linear anisotropic elasticity. The problem is governed by a second-order, linear, elliptic, partial differential equation with discontinuous coefficients. The most general anisotropy consistent with a state of anti-plane shear is considered, as well as a variety of boundary conditions. Anti-plane or longitudinal shear deformations are one of the simplest classes of deformations in solid mechanics. The resulting deformations are completely characterized by a single out-of-plane displacement which depends only on the in-plane coordinates. They can be thought of as complementary deformations to those of plane elasticity. While these deformations have received little attention compared with the plane problems of linear elasticity, they have recently been investigated for anisotropic and inhomogeneous linear elasticity. In the context of linear elasticity, Saint-Venant's principle is used to show that self-equilibrated loads generate local stress effects that quickly decay away from the loaded end of a structure. For homogeneous isotropic linear elastic materials this is well-documented. Self-equilibrated loads are a class of load distributions that are statically equivalent to zero, i.e., have zero resultant force and moment. When Saint-Venant's principle is valid, pointwise boundary conditions can be replaced by more tractable resultant conditions. It is shown in the present study that material inhomogeneity significantly affects the practical application of Saint-Venant's principle to sandwich structures.     

Remarks on ellipticity for the generalized Blatz-Ko constitutive model for a compressible nonlinearly elastic solid

One of the most widely used constitutive models for compressible isotropic nonlinearly elastic solids is the generalized Blatz-Ko material for foam-rubber and its various specializations. For this model, a unified derivation of necessary and sufficient conditions for ellipticity of the governing three-dimensional displacement equations of equilibrium is provided. When the parameterf occurring in the generalized Blatz-Ko model is in the range 0f<1, it is shown that ellipticity is always lost at sufficiently large stretches, while forf=1, the equilibrium equations are globally elliptic. The implications of these results for a variety of physical problems are discussed.     

A semi-inverse shape optimization problem in linear anti-plane shear

The design of a semi-infinite fillet for efficient stress transmission is considered. The problem is treated within the context of anti-plane shear deformations of a homogeneous, isotropic, linearly elastic solid. Under a remote state of simple shear, it is desired to determine the shape of the traction-free lateral boundaries of a symmetric plane domain so that the shear stress distribution on the finite end is as uniform as possible. A semi-inverse approach for a particular class of semi-infinite profiles is used to examine this issue.     

Solution of a Riemann problem for elasticity

This paper describes a numerical algorithm for the Riemann solution for nonlinear elasticity. We assume that the material is hyperelastic, which means that the stress-strain relations are given by the specific internal energy. Our results become more explicit under further assumptions: that the material is isotropic and that the Riemann problem is uniaxial. We assume that any umbilical points lie outside the region of physical relevance. Our main conclusion is that the Riemann solution can be obtained by the iterative solution of functional equations (Godunov iterations) each defined in one- or two-dimensional spaces.Supported in part by AFOSR-88-0025.     

Two illustrations of rapid crack growth in a pre-stressed compressible neo-Hookean material

An unbounded isotropic compressible neo-Hookean solid is initially in equilibrium under uniform tensile (possibly large) pre-stress. In one case, plane strain conditions generate slit crack growth at a constant sub-critical rate; in the other, axial symmetry produces penny-shaped crack growth. The procedure of superposing infinitesimal deformations upon those that are large is carried out in terms of tractable exact full-field solutions.These solutions are examined apart from a specific fracture mechanics model, nevertheless, they show that pre-stress induces, in addition to the expected anisotropy, a critical value above which a negative Poisson effect occurs. It is also found that dilatational, rotational and Rayleigh wave speeds decrease, and that the decrease is greater for the plane strain state associated with slit crack growth than for the axially symmetric state of the penny-shaped crack.Dynamic stress intensity factors are also extracted, and found to fall below those for a linear isotropic solid at the same pre-stress and crack growth rate. Moreover, the range of growth rates for sub-critical crack propagation is also decreased.     

Well-posedness of the equations of generalized thermoelasticity

In this paper the local existence, uniqueness and continuous dependence for smooth solutions to the initial value problem for a class of generalized (dependent on the time derivative of temperature) thermoelastic materials is proved. The field equations are written as a quasilinear hyperbolic system and the known results by Hughes, Kato and Marsden are applied.     

Uniqueness and continuous dependence of solutions to a multidimensional thermoelastic contact problem

Uniqueness and continuous dependence on the initial temperature are established for the solution of a multidimensional, quasistatic thermoelastic contact problem. The proof of this result does not depend on the ability to decouple the system of governing equations as required in the technique used by Shi and Shillor [European J. Appl. Math., 1990, 371–387] in the one dimensional analogue of this problem. Some extensions to other contact problems are suggested.     

On the Saint-Venant principle in the case of infinite energy

Estimates on the distribution of the elastic energy in a cylindrical domain in the context of linear elasticity are obtained. The estimates remain valid when the total elastic energy is infinite, and they can be used to establish Saint-Venant's principle without an assumption about finiteness of the total energy.Examples of boundary conditions resulting in infinite energy are constructed in the context of both linear elastostatics and special finite elastostatics, where a quadratic strain energy density function is assumed. The examples show that estimates of the type obtained are sometimes necessary.The results obtained are valid with obvious modifications in a space of any dimension n2.The results in this paper represent a partial fulfilment of the requirements for the degree of Doctor of Philosophy at Tel Aviv University by Y.S. under the guidance of J.J.R.     

A properly invariant theory of small deformations superposed on large deformations of an elastic rod

In the context of the direct or Cosserat theory of rods developed by Green, Naghdi and several of their co-workers, this paper is concerned with the development of a theory of small deformations which are superposed on large deformations. The resulting theory is properly invariant under all superposed rigid body motions. Furthermore, it is also valid for elastic rods which are subject to kinematical constraints, and it specializes to a linear theory of an elastic rod which is invariant under superposed rigid body motions. The construction of these theories is based on the method developed by Casey & Naghdi [1] who established similar theories for unconstrained nonpolar elastic bodies.     

Saint-Venant's problem for linear piezoelectric bodies

We seek a solution for a piezoelectric cylinder acted on the end faces by applied tractions and charges, under the hypothesis that both the stress and electric displacement fields depend linearly on the axial coordinate. The analysis, restricted to monoclinic materials of crystallographic class 2, leads to an explicit solution in terms of the strain and electric fields, which depend on the stress and charge resultants and on two scalar functions determined by the solution of a plane piezoelectric problem.     

A family of solutions describing plane strain cylindrical inflation in finite compressible elasticity

Within the context of finite, compressible, isotropic elasticity, a family of solutions describing plane strain cylindrical inflation of cylindrical shells is obtained for a class of materials that includes both the harmonic and Varga materials. Additionally it is shown that the class of materials chsen is the largest class of materials for which the family of solutions is possible.     

On minimal representations for constitutive equations of anisotropic elastic materials

The problem of determining minimal representations for anisotropic elastic constitutive equations is proposed and investigated. For elastic constitutive equations in any given case of anisotropy, it is shown that there exist generating sets consisting of six generators and such generating sets are minimal in all possible generating sets. This fact implies that most of the established results for representations of elastic constitutive equations are not minimal and remain to be sharpened. For elastic constitutive equations in some cases of anisotropy, including orthotropy, transverse isotropy, the trigonal crystal class S ₆, and the classes C _2mh , m=1, 2, 3,..., etc., representations in terms of minimal generating sets are presented for the first time.     

A reformulation of the strong ellipticity conditions for unconstrained hyperelastic media

The conditions for the strong ellipticity of the equilibrium equations of compressible, isotropic, nonlinearly elastic solids (established by Simpson and Spector [1]) are expressed in terms of the stored-energy function regarded as a function of the principal stretches. The applicability of this reformulation is illustrated with the help of two specific examples.     

Asymptotic analysis of linearly elastic shells: ‘Generalized membrane shells’

We consider a family of linearly elastic shells indexed by their half-thickness , all having the same middle surface % MathType!MTEF!2!1!+-% feaafiart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXafv3ySLgzGmvETj2BSbqefm0B1jxALjhiov2D% aebbfv3ySLgzGueE0jxyaibaiGc9yrFr0xXdbba91rFfpec8Eeeu0x% Xdbba9frFj0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs% 0dXdbPYxe9vr0-vr0-vqpWqaaeaabiGaciaacaqabeaadaqaaqGaaO% qaaiaadofacqGH9aqpcqaHvpGAcaGGOaGafqyYdCNbaebacaGGPaaa% aa!4317!\[S = \varphi (\bar \omega )\], with % MathType!MTEF!2!1!+-% feaafiart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXafv3ySLgzGmvETj2BSbqefm0B1jxALjhiov2D% aebbfv3ySLgzGueE0jxyaibaiGc9yrFr0xXdbba91rFfpec8Eeeu0x% Xdbba9frFj0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs% 0dXdbPYxe9vr0-vr0-vqpWqaaeaabiGaciaacaqabeaadaqaaqGaaO% qaaiabew9aQjaacQdacuaHjpWDgaqeaiabgkOimlaadkfadaahaaWc% beqaaiaaikdaaaGccqGHsgIRcaWGsbWaaWbaaSqabeaacaaIZaaaaa% aa!4812!\[\varphi :\bar \omega \subset R^2 \to R^3 \], and clamped along a portion of their lateral face whose trace on S is % MathType!MTEF!2!1!+-% feaafiart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXafv3ySLgzGmvETj2BSbqefm0B1jxALjhiov2D% aebbfv3ySLgzGueE0jxyaibaiGc9yrFr0xXdbba91rFfpec8Eeeu0x% Xdbba9frFj0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs% 0dXdbPYxe9vr0-vr0-vqpWqaaeaabiGaciaacaqabeaadaqaaqGaaO% qaaiabew9aQjaacIcacqaHZoWzdaWgaaWcbaGaaGimaaqabaGccaGG% Paaaaa!41EB!\[\varphi (\gamma _0 )\], where % MathType!MTEF!2!1!+-% feaafiart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXafv3ySLgzGmvETj2BSbqefm0B1jxALjhiov2D% aebbfv3ySLgzGueE0jxyaibaiGc9yrFr0xXdbba91rFfpec8Eeeu0x% Xdbba9frFj0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs% 0dXdbPYxe9vr0-vr0-vqpWqaaeaabiGaciaacaqabeaadaqaaqGaaO% qaaiaacIcacqaHZoWzdaWgaaWcbaGaaGimaaqabaGccaGGPaaaaa!401F!\[(\gamma _0 )\] is a fixed portion of with length % MathType!MTEF!2!1!+-% feaafiart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXafv3ySLgzGmvETj2BSbqefm0B1jxALjhiov2D% aebbfv3ySLgzGueE0jxyaibaiGc9yrFr0xXdbba91rFfpec8Eeeu0x% Xdbba9frFj0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs% 0dXdbPYxe9vr0-vr0-vqpWqaaeaabiGaciaacaqabeaadaqaaqGaaO% qaaiaacIcacqaHZoWzdaWgaaWcbaGaaGimaaqabaGccaGGPaGaeyOp% a4JaaGimaaaa!41E1!\[(\gamma _0 ) > 0\]. Let % MathType!MTEF!2!1!+-% feaafiart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXafv3ySLgzGmvETj2BSbqefm0B1jxALjhiov2D% aebbfv3ySLgzGueE0jxyaibaiGc9yrFr0xXdbba91rFfpec8Eeeu0x% Xdbba9frFj0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs% 0dXdbPYxe9vr0-vr0-vqpWqaaeaabiGaciaacaqabeaadaqaaqGaaO% qaaiaacIcacqaHZoWzdaWgaaWcbaGaeqySdeMaeqOSdigabeaakiaa% cIcacqaH3oaAcaGGPaGaaiykaaaa!45AA!\[(\gamma _{\alpha \beta } (\eta ))\] be the linearized strain tensor of S. We make an essential geometric and kinematic assumption, according to which the semi-norm % MathType!MTEF!2!1!+-% feaafiart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXafv3ySLgzGmvETj2BSbqefm0B1jxALjhiov2D% aebbfv3ySLgzGueE0jxyaibaiGc9yrFr0xXdbba91rFfpec8Eeeu0x% Xdbba9frFj0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs% 0dXdbPYxe9vr0-vr0-vqpWqaaeaabiGaciaacaqabeaadaqaaqGaaO% qaaiaacYhacqGHflY1caGG8bWaa0baaSqaaiabeM8a3bqaaiaad2ea% aaaaaa!4345!\[| \cdot |_\omega ^M \] defined by % MathType!MTEF!2!1!+-% feaafiart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXafv3ySLgzGmvETj2BSbqefm0B1jxALjhiov2D% aebbfv3ySLgzGueE0jxyaibaiGc9yrFr0xXdbba91rFfpec8Eeeu0x% Xdbba9frFj0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs% 0dXdbPYxe9vr0-vr0-vqpWqaaeaabiGaciaacaqabeaadaqaaqGaaO% qaaiaacYhacqaH3oaAcaGG8bWaa0baaSqaaiabeM8a3bqaaiaad2ea% aaGccqGH9aqpdaGadeqaamaaqababaGaaiiFaiaacYhaaSqaaiabeg% 7aHfrbbjxAHXgaiuaacaWFSaGaeqOSdigabeqdcqGHris5aOGaeq4S% dCMaeqySdeMaeqOSdiMaaiikaiabeE7aOjaacMcacaGG8bGaaiiFam% aaDaaaleaacaWGmbWaaWbaaWqabeaacaaIYaaaaSGaaiikaiabeM8a% 3jaacMcaaeaacaaIYaaaaaGccaGL7bGaayzFaaWaaWbaaSqabeaaca% aIXaGaai4laiaaikdaaaaaaa!61F1!\[|\eta |_\omega ^M = \left\{ {\sum\nolimits_{\alpha ,\beta } {||} \gamma \alpha \beta (\eta )||_{L^2 (\omega )}^2 } \right\}^{1/2} \] is a norm over the space % MathType!MTEF!2!1!+-% feaafiart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXafv3ySLgzGmvETj2BSbqefm0B1jxALjhiov2D% aebbfv3ySLgzGueE0jxyaibaiGc9yrFr0xXdbba91rFfpec8Eeeu0x% Xdbba9frFj0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs% 0dXdbPYxe9vr0-vr0-vqpWqaaeaabiGaciaacaqabeaadaqaaqGaaO% qaaiaadAfacaGGOaGaeqyYdCNaaiykaiabg2da9maacmqabaGaeq4T% dGMaeyicI4SaamisamaaCaaaleqabaGaaGymaaaakiaacIcacqaHjp% WDcaGGPaGaai4oaiabeE7aOjabg2da9iaab+gacaqGUbGaeq4SdC2a% aSbaaSqaaiaabcdaaeqaaaGccaGL7bGaayzFaaaaaa!5361!\[V(\omega ) = \left\{ {\eta \in H^1 (\omega );\eta = {\text{on}}\gamma _{\text{0}} } \right\}\], excluding however the already analyzed membrane shells, where % MathType!MTEF!2!1!+-% feaafiart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXafv3ySLgzGmvETj2BSbqefm0B1jxALjhiov2D% aebbfv3ySLgzGueE0jxyaibaiGc9yrFr0xXdbba91rFfpec8Eeeu0x% Xdbba9frFj0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs% 0dXdbPYxe9vr0-vr0-vqpWqaaeaabiGaciaacaqabeaadaqaaqGaaO% qaaiabeo7aNnaaBaaaleaacaqGWaaabeaakiabg2da9iabgkGi2kab% eM8a3baa!42F8!\[\gamma _{\text{0}} = \partial \omega \] and S is elliptic. This new assumption is satisfied for instance if % MathType!MTEF!2!1!+-% feaafiart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXafv3ySLgzGmvETj2BSbqefm0B1jxALjhiov2D% aebbfv3ySLgzGueE0jxyaibaiGc9yrFr0xXdbba91rFfpec8Eeeu0x% Xdbba9frFj0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs% 0dXdbPYxe9vr0-vr0-vqpWqaaeaabiGaciaacaqabeaadaqaaqGaaO% qaaiabeo7aNnaaBaaaleaacaqGWaaabeaakiabgcMi5kabgkGi2kab% eM8a3baa!43B9!\[\gamma _{\text{0}} \ne \partial \omega \] and S is elliptic, or if S is a portion of a hyperboloid of revolution.We then show that, as 0, the averages % MathType!MTEF!2!1!+-% feaafiart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXafv3ySLgzGmvETj2BSbqefm0B1jxALjhiov2D% aebbfv3ySLgzGueE0jxyaibaiGc9yrFr0xXdbba91rFfpec8Eeeu0x% Xdbba9frFj0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs% 0dXdbPYxe9vr0-vr0-vqpWqaaeaabiGaciaacaqabeaadaqaaqGaaO% qaamaalaaabaGaaGymaaqaaiaaikdadaahaaWcbeqaaiabew7aLbaa% aaGcdaWdXaqaaiaadwhadaqhaaWcbaGaamyAaaqaaiabew7aLbaaki% aabsgacaWG4bWaa0baaSqaaiaaiodaaeaacqaH1oqzaaaabaGaeyOe% I0IaeqyTdugabaGaeqyTduganiabgUIiYdaaaa!4E28!\[\frac{1}{{2^\varepsilon }}\int_{ - \varepsilon }^\varepsilon {u_i^\varepsilon {\text{d}}x_3^\varepsilon } \] across the thickness of the shell of the covariant components u _i of the displacement of the points of the shell strongly converge in the completion V _M ^#() of V() with respect to the norm % MathType!MTEF!2!1!+-% feaafiart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXafv3ySLgzGmvETj2BSbqefm0B1jxALjhiov2D% aebbfv3ySLgzGueE0jxyaibaiGc9yrFr0xXdbba91rFfpec8Eeeu0x% Xdbba9frFj0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs% 0dXdbPYxe9vr0-vr0-vqpWqaaeaabiGaciaacaqabeaadaqaaqGaaO% qaaiaacYhacqGHflY1caGG8bWaa0baaSqaaiabeM8a3bqaaiaad2ea% aaaaaa!4345!\[| \cdot |_\omega ^M \], toward the solution of a generalized membrane shell problem. This convergence result also justifies the recent formal asymptotic approach of D. Caillerie and E. Sanchez-Palencia.The limit problem found in this fashion is sensitive, according to the terminology recently introduced by J.L. Lions and E. Sanchez-Palencia, in the sense that it possesses two unusual features: it is posed in a space that is not necessarily contained in a space of distributions, and its solution is highly sensitive to arbitrarily small smooth perturbations of the data.Under the same assumption, we also show that the average % MathType!MTEF!2!1!+-% feaafiart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXafv3ySLgzGmvETj2BSbqefm0B1jxALjhiov2D% aebbfv3ySLgzGueE0jxyaibaiGc9yrFr0xXdbba91rFfpec8Eeeu0x% Xdbba9frFj0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs% 0dXdbPYxe9vr0-vr0-vqpWqaaeaabiGaciaacaqabeaadaqaaqGaaO% qaamaalaaabaGaaGymaaqaaiaaikdadaahaaWcbeqaaiabew7aLbaa% aaGcdaWdXaqaaiaadwhadaqhaaWcbaGaamyAaaqaaiabew7aLbaaki% aabsgacaWG4bWaa0baaSqaaiaaiodaaeaacqaH1oqzaaaabaGaeyOe% I0IaeqyTdugabaGaeqyTduganiabgUIiYdaaaa!4E28!\[\frac{1}{{2^\varepsilon }}\int_{ - \varepsilon }^\varepsilon {u_i^\varepsilon {\text{d}}x_3^\varepsilon } \] where % MathType!MTEF!2!1!+-% feaafiart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXafv3ySLgzGmvETj2BSbqefm0B1jxALjhiov2D% aebbfv3ySLgzGueE0jxyaibaiGc9yrFr0xXdbba91rFfpec8Eeeu0x% Xdbba9frFj0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs% 0dXdbPYxe9vr0-vr0-vqpWqaaeaabiGaciaacaqabeaadaqaaqGaaO% qaaiaadwhadaahaaWcbeqaaiabew7aLbaakiabg2da9iaacIcacaWG% 1bWaa0baaSqaaiaadMgaaeaacqaH1oqzaaGccaGGPaaaaa!452C!\[u^\varepsilon = (u_i^\varepsilon )\], and the solution % MathType!MTEF!2!1!+-% feaafiart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXafv3ySLgzGmvETj2BSbqefm0B1jxALjhiov2D% aebbfv3ySLgzGueE0jxyaibaiGc9yrFr0xXdbba91rFfpec8Eeeu0x% Xdbba9frFj0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs% 0dXdbPYxe9vr0-vr0-vqpWqaaeaabiGaciaacaqabeaadaqaaqGaaO% qaaiabe67a4naaCaaaleqabaGaeqyTdugaaOGaeyicI4SaamOvamaa% BaaaleaacaWGlbaabeaakiaacIcacqaHjpWDcaGGPaaaaa!465B!\[\xi ^\varepsilon \in V_K (\omega )\] of Koiter's equations have the same principal part as 0 in the same space V _M () as above. For such generalized membrane shells, the two-dimensional shell model of W.T. Koiter is thus likewise justified.We also treat the case where % MathType!MTEF!2!1!+-% feaafiart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXafv3ySLgzGmvETj2BSbqefm0B1jxALjhiov2D% aebbfv3ySLgzGueE0jxyaibaiGc9yrFr0xXdbba91rFfpec8Eeeu0x% Xdbba9frFj0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs% 0dXdbPYxe9vr0-vr0-vqpWqaaeaabiGaciaacaqabeaadaqaaqGaaO% qaaiaacYhacqGHflY1caGG8bWaa0baaSqaaiabeM8a3bqaaiaad2ea% aaaaaa!4345!\[| \cdot |_\omega ^M \] is no longer a norm over V(), but is a norm over the space % MathType!MTEF!2!1!+-% feaafiart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXafv3ySLgzGmvETj2BSbqefm0B1jxALjhiov2D% aebbfv3ySLgzGueE0jxyaibaiGc9yrFr0xXdbba91rFfpec8Eeeu0x% Xdbba9frFj0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs% 0dXdbPYxe9vr0-vr0-vqpWqaaeaabiGaciaacaqabeaadaqaaqGaaO% qaaiaadAfadaWgaaWcbaGaam4saaqabaGccaGGOaGaeqyYdCNaaiyk% aiabg2da9maacmqabaGaeq4TdGMaeyypa0JaaiikaiabeE7aOnaaBa% aaleaacaWGPbaabeaakiaacMcacqGHiiIZcaWGibWaaWbaaSqabeaa% caaIXaaaaOGaaiikaiabeM8a3jaacMcacqGHxdaTcaWGibWaaWbaaS% qabeaacaaIYaaaaOGaaiikaiabeM8a3jaacMcacaGG7aGaeq4TdG2a% aSbaaSqaaiaadMgaaeqaaOGaeyypa0JaeyOaIy7aaSbaaSqaaiaadA% haaeqaaOGaeq4TdG2aaSbaaSqaaiaaiodaaeqaaOGaeyypa0JaaGim% aiGac+gacaGGUbGaeq4SdC2aaSbaaSqaaiaaicdaaeqaaaGccaGL7b% GaayzFaaaaaa!68B8!\[V_K (\omega ) = \left\{ {\eta = (\eta _i ) \in H^1 (\omega ) \times H^2 (\omega );\eta _i = \partial _v \eta _3 = 0\operatorname{on} \gamma _0 } \right\}\], thus also excluding the already analyzed flexural shells. Then a convergence theorem can still be established, but only in the completion of the quotient space V()/V ₀() with repect to % MathType!MTEF!2!1!+-% feaafiart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXafv3ySLgzGmvETj2BSbqefm0B1jxALjhiov2D% aebbfv3ySLgzGueE0jxyaibaiGc9yrFr0xXdbba91rFfpec8Eeeu0x% Xdbba9frFj0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs% 0dXdbPYxe9vr0-vr0-vqpWqaaeaabiGaciaacaqabeaadaqaaqGaaO% qaaiaacYhacqGHflY1caGG8bWaa0baaSqaaiabeM8a3bqaaiaad2ea% aaaaaa!4345!\[| \cdot |_\omega ^M \], where % MathType!MTEF!2!1!+-% feaafiart1ev1aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn% hiov2DGi1BTfMBaeXafv3ySLgzGmvETj2BSbqefm0B1jxALjhiov2D% aebbfv3ySLgzGueE0jxyaibaiGc9yrFr0xXdbba91rFfpec8Eeeu0x% Xdbba9frFj0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs% 0dXdbPYxe9vr0-vr0-vqpWqaaeaabiGaciaacaqabeaadaqaaqGaaO% qaaiaadAfadaWgaaWcbaGaaGimaaqabaGccaGGOaGaeqyYdCNaaiyk% aiabg2da9maacmqabaGaeq4TdGMaeyicI4SaamOvaiaacIcacqaHjp% WDcaGGPaGaai4oaiabeo7aNjabeg7aHjabek7aIjaacIcacqaH3oaA% caGGPaGaeyypa0JaaeimaiaabMgacaqGUbGaeqyYdChacaGL7bGaay% zFaaaaaa!5997!\[V_0 (\omega ) = \left\{ {\eta \in V(\omega );\gamma \alpha \beta (\eta ) = {\text{0in}}\omega } \right\}\].These convergence results, together with those that we already obtained for membrane and flexural shells, jointly with B. Miara in the second case, thus constitute an asymptotic analysis of linearly elastic shells in all possible cases.     

Saint-Venant's problem with Voigt's hypotheses for anisotropic solids

We seek for a solution of Saint-Venant's problem for inhomogeneous and anisotropic materials under the assumptions, introduced by Voigt, that the stress is either constant along the axis of the cylinder or depends linearly on the axial coordinate. We first prove the uniqueness of the solution in terms of resultants, then we exhibit an explicit formula for such a solution; we show finally how Clebsch's hypothesis, that the stress vector on axial planes is parallel to the axis, is compatible with Voigt's hypotheses provided that the symmetry group of the material comprising the cylinder contains the reflections on the cross-section.     

The tensor equation AX+XA=Φ(A,H), with applications to kinematics of continua

The (second-order) tensor equation AX+XA=(A,H) is studied for certain isotropic functions (A,H) which are linear in H. Qualitative properties of the solution X and relations between the solutions for various forms of are established for an inner product space of arbitrary dimension. These results, together with Rivlin's identities for tensor polynomials in two variables, are applied in three dimensions to obtain new explicit formulas for X in direct tensor notation as well as new derivations of previously known formulas. Several applications to the kinematics of continua are considered.     

Existence of Weak Solutions for the Unsteady Interaction of a Viscous Fluid with an Elastic Plate

The purpose of this work is to study the existence of solutions for an unsteady fluid-structure interaction problem. We consider a three-dimensional viscous incompressible fluid governed by the Navier–Stokes equations, interacting with a flexible elastic plate located on one part of the fluid boundary. The fluid domain evolves according to the structure’s displacement, itself resulting from the fluid force. We prove the existence of at least one weak solution as long as the structure does not touch the fixed part of the fluid boundary. The same result holds also for a two-dimensional fluid interacting with a one-dimensional membrane.     

Divergence-Free Solutions of Poisson-Like Equations

It is shown that Poisson-like equations for 3-vector fields can have divergence-free solutions and desired boundary behaviour in bounded domains, provided the mean curvature of the boundary is nowhere too large positive. A-priori estimates for the solutions are given.