The traction problem in plane incompressible linear elastostatics in exterior domains (Q1368503)

Let \(B=\mathbb{R}^2 \setminus B_0\) be a connected set, where \(B_0\) is a convex and closed set, \(\partial B\) of class \(C^2\), \(b\in L^1(B, \mathbb{R}^2)\cap C(B, \mathbb{R}^2)\), \[ \int_B \bigl(| x| \ln| x| \bigr)^2 \bigl| b(x)\bigr|^2dx< +\infty,\;s\in C(\partial B, \mathbb{R}^2), \quad\int_B b(x)dx+ \int_{\partial B} s(x)d \sigma=0. \] Lin and Sym denote the space of linear maps from \(\mathbb{R}^2\) to \(\mathbb{R}^2\) and the space of linear symmetric maps from \(\mathbb{R}^2\) to \(\mathbb{R}^2\) respectively; for all \(x\in B\) or \(x=\infty\) let \(C(x): \text{Lin} \to\text{Sym}\) be a linear map such that \(C(x)(W)=0\) for every linear skew map \(W\) from \(\mathbb{R}^2\) to \(\mathbb{R}^2\) and there exist \(m,M>0\) such that \(m| L|^2\leq L\cdot C(x) (L)\leq M| L |^2\) for all \(L\in\text{Sym}\), we suppose \(C\) is continuously differentiable on \(\overline B\). The author proves the following results: Theorem 1. There exists \(u\in F\equiv \{\varphi \in L^2_{\text{loc}} (B,\lambda^2): \widehat\nabla \varphi\in L^2_{\text{loc}} (B),\text{div} \varphi =0\}\), where \(\widehat\nabla \varphi\) is the symmetric part of \(\nabla\varphi\), such that \[ \int_B\nabla \varphi(x) \cdot C(x)\bigl(\nabla u(x)\bigr)dx =\int_B\bigl(\varphi (x)-W(x)\bigr) \cdot b(x)dx+ \int_{\partial B} \bigl(\varphi (x)-W(x)\bigr) \cdot s(x)d \sigma \] for every \(\varphi\in F\), where \[ W=\lim_{R\to +\infty} {1\over|\{x \in B:| x|<R\} |} \int_{\{x\in B:| x|<R\}} \widetilde\nabla \varphi(x)dx \] and \(\widetilde \nabla \varphi\) is the skew part of \(\nabla\varphi\); moreover \[ \int_B\bigl| \widehat\nabla u(x) \bigr |^2 dx\leq c\left(\int_B \bigl(| x| \ln| x|\bigr)^2 \bigl| b(x)\bigr|^2 dx+\int_{ \partial B} \bigl| s(x)\bigr|^2 d\sigma\right) \text{ for } c>0. \] Theorem 2. Let \(u\in F\) such that \(\int_B\nabla \varphi(x) \cdot C(x)(\nabla u(x)) dx=0\) and assume that one of the following conditions is satisfied: (i) \(\int_{\{x\in B:R<| x|<2R\}} |\widehat \nabla u(x)|^2 dx=o(R^\alpha)\), where \[ \alpha= {1-\xi_1 \over\bigl( 1+\sqrt 2\bigr) \sqrt{c{M \over m}} +2\xi} \text{ for symmetric } C,\quad \alpha= {1-\xi_1\over \bigl(1+\sqrt 2\bigr) \sqrt c{M \over m} +2\xi} \text{ otherwise}, \] for \(\xi\), \(\xi_1>0\); (ii) \(C\) is constant and \(\int_{\{x\in B:R< | x| <2R\}}| \widehat\nabla u(x)|^2 dx=O (R^{2-\varepsilon})\) for \(\varepsilon>0\); (iii) there exists \(\mu_0>0\) such that \(\mu_0| A|^2\leq A\bullet C(\infty) (A)\), \(| C(x)-C(\infty) |\leq \mu_0 \delta\) for all \(| x|\geq R_0\) and for a \(\delta\in ]0,1[\) and \(\int_{\{x\in B:R< | x|<2R\}} |\widehat \nabla u(x)|^2 dx=o(R^\zeta)\), where \(\zeta \in]0, 2[\) is such that \(16\delta^2 <(\zeta/72 c)^{\zeta/(2-\zeta)} (1-\zeta/2)\); then \(\int_B| \widehat \nabla u(x)|^2 dx=0\).