Asymptotic behavior of a nonlinear shallow shell model when the shell becomes a plate

Trung Hieu Giang, Ngoc Quynh Nguyen · 2025

Read Asymptotic behavior of a nonlinear shallow shell model when the shell becomes a plate by Trung Hieu Giang, Ngoc Quynh Nguyen (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-22677.txt.

Introduction

Trung Hieu Giang and Ngoc Quynh Nguyen study what happens to a curved-shell model as the shell flattens into a plate. Their subject is the nonlinear shallow shell model associated with Donnell, Vlasov, Mushtari, Galimov, and Koiter, a classic description of thin curved forms under load. The question is asymptotic: do minimizing solutions of the shell model settle into the solutions of the corresponding plate model as curvature disappears?

The authors prove that they do, and their result allows general applied forces. That generality matters because the earlier reference result assumed the tangential components of the applied forces vanish, a key restriction. Removing it substantially widens the class of loading conditions under which the shell-to-plate limit is now justified.

Catalog record

CREATORTrung Hieu Giang, Ngoc Quynh Nguyen
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE40 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2512-22677

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Asymptotic behavior of a nonlinear shallow shell model when the shell becomes a plate Trung Hieu Giang, Ngoc Quynh Nguyen Abstract This paper studies a nonlinear shallow shell model proposed by Donnell, Vlasov, Mushtari, Galimov, and Koiter. More specifically, we address the question concerning the asymptotic behavior of minimizing solutions. Our result can be applied to general applied forces. Thus, it substantially extends the one given in [ 13 ] whereby the tangential components of the applied forces are assumed to vanish. 2020 Mathematics Subject Classification: 74B20, 35B40, 74K25. Keywords: nonlinear elasticity, asymptotic behavior, equilibrium problems, shallow shells. 1 Notation Throughout this paper, Greek indices and exponents range in the set \{1,2\} , while Latin indices and exponents range in the set \{1,2,3\} , unless they are used for indexing sequences or otherwise specified in the text. The summation convention with respect to repeated indices and exponents is also used. Strong and weak convergences in any normed vector space are respectively denoted \to\text{ and }\rightharpoonup. The notation y=(y_{\alpha}) denotes a generic point in \mathbb{R}^{2} and partial derivatives, in the classical or distributional sense, are denoted \partial_{\alpha}:=\partial/\partial y_{\alpha} and \partial_{\alpha\beta}:=\partial^{2}/\partial y_{\alpha}\partial y_{\beta} . Vector and matrix fields are denoted by boldface letters. The Euclidean norm, the inner product, and the vector product of two vectors \boldsymbol{u} and \boldsymbol{v} in \mathbb{R}^{3} are respectively denoted |\boldsymbol{u}| , \boldsymbol{u}\cdot\boldsymbol{v} , and \boldsymbol{u}\wedge\boldsymbol{v} . The notation (c_{\alpha\beta}) designates the 2\times 2 matrix whose component at its \alpha -row and \beta -column is c_{\alpha\beta} . The usual norm of the Lebesgue space L^{p}(\omega) , 1\leq p\leq\infty , is denoted by \|\cdot\|_{L^{p}(\omega)} . The notation L^{p}(\omega;\mathbb{R}^{m\times n}) denotes, for every positive integer m and n , the space of matrix fields \boldsymbol{A}=(A_{ij}):\omega\to\mathbb{R}^{m\times n} with components A_{ij} in the Lebesgue space L^{p}(\omega) . This space of matrix-valued functions is equipped with the norm \|\boldsymbol{A}\|_{L^{p}(\omega)}:=\Big(\sum\limits_{i=1}^{m}\sum\limits_{j=1}^{n}\|A_{ij}\|^{p}_{L^{p}(\omega)}\Big)^{1/p}\textrm{ for all }\boldsymbol{A}\in L^{p}(\omega;\mathbb{R}^{m\times n}). Note that we use the same notation for the norm of scalar, vector, and matrix-valued functions. The usual norm of the Sobolev space W^{m,p}(\omega) , m\in\mathbb{N}^{*} , 1\leq p\leq\infty , is denoted by \|\cdot\|_{W^{m,p}(\omega)} . The notations H^{1}(\omega) and H^{2}(\omega) respectively denote the spaces W^{1,2}(\omega) and W^{2,2}(\omega) . We also denote by H^{1}_{0}(\omega) and H^{2}_{0}(\omega) respectively the closures of \mathcal{D}(\omega) in H^{1}(\omega) and H^{2}(\omega) . In this paper, by shell we mean a three-dimensional body whose natural state (that is, a stress-free configuration of the body) is a “surface with positive thickness” in the three-dimensional Euclidean space. This means that a shell is a subset of {\mathbb{R}}^{3} of the form \overline{\boldsymbol{\Theta}(\Omega)}, where \Omega:=\omega\times(-\varepsilon,\varepsilon) and \boldsymbol{\Theta}(y,x_{3}):=\boldsymbol{\theta}(y)+x_{3}\boldsymbol{a}_{3,\boldsymbol{\theta}}(y),\ (y,x_{3})\in\Omega, for some given bounded and connected open set \omega\subset{\mathbb{R}}^{2} whose boundary is Lipschitz-continuous in the sense of Adams & Fournier [ 1 ] , real number 0<\varepsilon\leq\varepsilon_{0} , and immersion \boldsymbol{\theta}:\overline{\omega}\to\mathbb{R}^{3} of class {\mathcal{C}}^{2} with a unit vector field \boldsymbol{a}_{3,\boldsymbol{\theta}}:\overline{\omega}\to\mathbb{R}^{3} defined by \boldsymbol{a}_{3,\boldsymbol{\theta}}(y):=\dfrac{{\partial}_{1}\boldsymbol{\theta}(y)\wedge{\partial}_{2}\boldsymbol{\theta}(y)}{\left|{\partial}_{1}\boldsymbol{\theta}(y)\wedge{\partial}_{2}\boldsymbol{\theta}(y)\right|}\textrm{ for all }y\in\overline{\omega}. (1) The assumption that \boldsymbol{\theta} is an immersion of class {\mathcal{C}}^{2} means that \boldsymbol{\theta}\in{\mathcal{C}}^{2}(\omega;{\mathbb{R}}^{3}) and the two vector fields \boldsymbol{a}_{\alpha,\boldsymbol{\theta}}:={\partial}_{\alpha}\boldsymbol{\theta} are linearly independent at every y\in\omega , so that the unit normal vector field \boldsymbol{a}_{3,\boldsymbol{\theta}} in ( 1 ) is well defined and of class \mathcal{C}^{1} . The real number \varepsilon>0 represents the half-thickness of the shell, and S:=\boldsymbol{\theta}(\overline{\omega}) represents the middle surface of the shell. The upper bound \varepsilon_{0} of \varepsilon is chosen sufficiently small compared with the diameter of \omega , so that \boldsymbol{\Theta} is itself an immersion of class {\mathcal{C}}^{1} from \overline{\Omega} into \mathbb{R}^{3} (the existence of such a small \varepsilon_{0} is proved in [ 2 ] ). Note that \boldsymbol{a}_{3,\boldsymbol{\theta}}(y) is the positively oriented unit vector normal to the surface S at the point \boldsymbol{\theta}(y) , while the two vectors \boldsymbol{a}_{\alpha,\boldsymbol{\theta}}(y) form a basis in the tangent plane to S at the same point. The three vectors \boldsymbol{a}_{i,\boldsymbol{\theta}}(y) constitute the covariant basis in \mathbb{R}^{3} at the point \boldsymbol{\theta}(y) , while the three vectors \boldsymbol{a}^{i}_{\boldsymbol{\theta}}(y) uniquely defined by the relations \boldsymbol{a}^{i}_{\boldsymbol{\theta}}(y)\cdot\boldsymbol{a}_{j,\boldsymbol{\theta}}(y)=\delta^{i}_{j}, where \delta^{i}_{j} is the Kronecker symbol, constitute the contravariant basis at the point \boldsymbol{\theta}(y)\in S . As a consequence, any vector field \boldsymbol{\zeta}:\omega\to\mathbb{R}^{3} can be decomposed over either of these bases as \boldsymbol{\zeta}=\zeta_{i}\boldsymbol{a}^{i}_{\boldsymbol{\theta}}=\zeta^{i}\boldsymbol{a}_{i,\boldsymbol{\theta}}, for some functions \zeta_{i}:\omega\to\mathbb{R} and \zeta^{i}:\omega\to\mathbb{R} . The area element of the surface S is denoted by \sqrt{a_{\boldsymbol{\theta}}(y)}dy , where a_{\boldsymbol{\theta}}:=|\partial_{1}\boldsymbol{\theta}\wedge\partial_{2}\boldsymbol{\theta}|^{2}\text{ in }\omega. (2) The covariant components a_{\alpha\beta,\boldsymbol{\theta}}\in\mathcal{C}^{1}(\overline{\omega}) and b_{\alpha\beta,\boldsymbol{\theta}}\in\mathcal{C}^{0}(\overline{\omega}) of respectively the first and second fundamental forms of S=\boldsymbol{\theta}(\overline{\omega}) are defined by a_{\alpha\beta,\boldsymbol{\theta}}:=\boldsymbol{a}_{\alpha,\boldsymbol{\theta}}\cdot\boldsymbol{a}_{\beta,\boldsymbol{\theta}}\textrm{ and }b_{\alpha\beta,\boldsymbol{\theta}}:=\boldsymbol{a}_{3,\boldsymbol{\theta}}\cdot\partial_{\alpha\beta}\boldsymbol{\theta}\textrm{ in }\overline{\omega}. The contravariant components of the first fundamental form are the components a^{\alpha\beta}_{\boldsymbol{\theta}}\in\mathcal{C}^{1}(\overline{\omega}) of the inverse matrix (a^{\alpha\beta}_{\boldsymbol{\theta}}(y)):=(a_{\alpha\beta,\boldsymbol{\theta}}(y))^{-1},\ y\in\overline{\omega}. The Christoffel symbols \Gamma^{\sigma}_{\alpha\beta,\boldsymbol{\theta}}\in\mathcal{C}^{0}(\overline{\omega}) of S are defined by \Gamma^{\sigma}_{\alpha\beta,\boldsymbol{\theta}}:=\boldsymbol{a}^{\sigma}_{\boldsymbol{\theta}}\cdot\partial_{\beta}\boldsymbol{a}_{\alpha,\boldsymbol{\theta}}\textrm{ in }\overline{\omega}. The Gaussian curvature of the surface S=\boldsymbol{\theta}(\overline{\omega}) is defined by K_{\boldsymbol{\theta}}:=\det\left(a^{\alpha\sigma}_{\boldsymbol{\theta}}b_{\sigma\beta,\boldsymbol{\theta}}\right)\in{\mathcal{C}}^{0}(\overline{\omega}). 2 The nonlinear shallow shell model The lower-dimensional elasticity theories play an important role in mathematical analysis and numerical computations. On the one hand, they allow simpler computational approaches; on the other hand, they often possess a richer variety of mathematical results. In recent decades, formal asymptotic analysis, based on various additional mathematical and mechanical assumptions, has led to a significant number of different two-dimensional shell theories, such as the nonlinear membrane shell, flexural shell, Koiter shell, etc. (for a more complete list of shell theories, we refer to [ 4 ] and the references therein). However, to date, many theoretical aspects of these models are still not fully understood. This paper is concerned with a nonlinear shallow shell model, which is a two-dimensional theory proposed by L.H. Donnell, V.Z. Vlasov, K.M. Mushtari & K.Z. Galimov, and W.T. Koiter [ 8 , 10 , 11 , 15 ] . In this context, a shell is called “shallow” when the Gaussian curvature K_{\boldsymbol{\theta}} of its middle surface is sufficiently small. The nonlinear shallow shell model is applied to the shells made of an elastic material whose behavior is modeled by two elastic coefficients, denoted \lambda\geq 0 and \mu>0 , via the two-dimensional elasticity tensor (a_{\boldsymbol{\theta}}^{\alpha\beta\sigma\tau}) , where a_{\boldsymbol{\theta}}^{\alpha\beta\sigma\tau}:=\dfrac{4\lambda\mu}{\lambda+2\mu}a_{\boldsymbol{\theta}}^{\alpha\beta}a_{\boldsymbol{\theta}}^{\sigma\tau}+2\mu(a_{\boldsymbol{\theta}}^{\alpha\sigma}a_{\boldsymbol{\theta}}^{\beta\tau}+a_{\boldsymbol{\theta}}^{\alpha\tau}a_{\boldsymbol{\theta}}^{\beta\sigma}). (3) Notice that this tensor is positive-definite in the sense that there exists a constant c_{e}(\omega,\boldsymbol{\theta},\lambda,\mu)>0 such that c_{e}(\omega,\boldsymbol{\theta},\lambda,\mu)\sum\limits_{\alpha,\beta}|t_{\alpha\beta}|^{2}\leq a_{\boldsymbol{\theta}}^{\alpha\beta\sigma\tau}(y)t_{\sigma\tau}t_{\alpha\beta}, (4) for all y\in\overline{\omega} and all symmetric matrices (t_{\alpha\beta}) (see, e.g., [ 4 , Theorem 3.3-2] ). For simplicity, our analysis in this paper is restricted to the case of totally clamped shells. Then, the nonlinear shallow shell model is often formulated as a minimization problem as follows: Problem \mathcal{P}_{shallow,\boldsymbol{\theta}} . Find \boldsymbol{u}=(u_{i})\in\boldsymbol{V}(\omega) that minimizes the energy \displaystyle J_{\boldsymbol{\theta}}(\boldsymbol{\eta}):= \displaystyle\frac{1}{2}\int\limits_{\omega}\Big\{\frac{\varepsilon^{3}}{3}a_{\boldsymbol{\theta}}^{\alpha\beta\sigma\tau}F_{\sigma\tau}^{\boldsymbol{\theta}}(\boldsymbol{\eta})F_{\alpha\beta}^{\boldsymbol{\theta}}(\boldsymbol{\eta})+\varepsilon a_{\boldsymbol{\theta}}^{\alpha\beta\sigma\tau}E^{\boldsymbol{\theta}}_{\sigma\tau}(\boldsymbol{\eta})E^{\boldsymbol{\theta}}_{\alpha\beta}(\boldsymbol{\eta})\Big\}\sqrt{a_{\boldsymbol{\theta}}}\,dy \displaystyle-\int\limits_{\omega}p^{i}\eta_{i}\sqrt{a_{\boldsymbol{\theta}}}\,dy,\textrm{ for every }\boldsymbol{\eta}\in\boldsymbol{V}(\omega), (5) where \boldsymbol{V}(\omega):=H^{1}_{0}(\omega)\times H^{1}_{0}(\omega)\times H^{2}_{0}(\omega), E^{\boldsymbol{\theta}}_{\alpha\beta}(\boldsymbol{\eta}):=\dfrac{1}{2}\big(\partial_{\beta}\eta_{\alpha}+\partial_{\alpha}\eta_{\beta}\big)-\Gamma^{\sigma}_{\alpha\beta,\boldsymbol{\theta}}\eta_{\sigma}-b_{\alpha\beta,\boldsymbol{\theta}}\eta_{3}+\dfrac{1}{2}\partial_{\alpha}\eta_{3}\partial_{\beta}\eta_{3}, F^{\boldsymbol{\theta}}_{\alpha\beta}(\boldsymbol{\eta}):=\partial_{\alpha\beta}\eta_{3}-\Gamma^{\sigma}_{\alpha\beta,\boldsymbol{\theta}}\partial_{\sigma}\eta_{3}, and the functions p_{i}\in L^{2}(\omega) are the components of the density \boldsymbol{p}=(p_{i}) of the applied force. Here, each admissible field \boldsymbol{\eta}=(\eta_{i})\in\boldsymbol{V}(\omega) corresponds to an admissible displacement \eta_{i}\boldsymbol{a}^{i}_{\boldsymbol{\theta}} of the middle surface \boldsymbol{\theta}(\overline{\omega}) . Over the last several decades, this nonlinear shallow shell model has been studied by several authors, with a primary focus on the existence, uniqueness, and asymptotic behavior of solutions. Regarding the Euler-Lagrange equation associated with the functional J_{\boldsymbol{\theta}} , by using the theory of pseudo-monotone operators, Bernadou & Oden [ 5 ] have established the existence of a weak solution if the tangential components of the applied forces and the curvature of the middle surface are sufficiently small. They also proved that the weak solution found is unique if the applied forces are sufficiently small. Figueiredo [ 9 ]

Read the complete text