Exact boundary controllability for the ideal magneto-hydrodynamic equations (Q2670009)

The authors consider an exact controllability problem associated to the ideal magneto-hydrodynamic (MHD) equations written as: \(\partial _{t}u+u\cdot \nabla u-B\cdot \nabla B+\nabla p=0\), \(\partial _{t}B+u\cdot \nabla B-B\cdot \nabla u=0\), \(\operatorname{div}u=\operatorname{div}B=0\), and posed in \(\Omega =[0,1]^{2}\) , \(u\) being the velocity, \(B\) the magnetic field, and \(p\) the pressure. The authors take \(\Gamma =\{0,1\}\times \lbrack 0,1]\) as the controlled part of the boundary and the controllability problem is raised in the following way: for given \(T>0\), initial data \((u_{0},B_{0})\), and terminal data \( (u_{1},B_{1})\) satisfying \(\operatorname{div}u_{0}=\operatorname{div}u_{1}=\operatorname{div}B_{0}=\operatorname{div}B_{1}=0\) in \(\Omega \), \(u_{0}\cdot n=u_{1}\cdot n=B_{0}\cdot n=B_{1}\cdot n=0\) on \(\partial \Omega \setminus \Gamma \), does there exist a solution \((u,B)\) to the above MHD equations such that \((u,B)\mid _{t=0}=(u_{0},B_{0})\), \((u,B)\mid _{t=T}=(u_{1},B_{1})\), and \(u\cdot n=B\cdot n=0\), \(t\in \lbrack 0,T]\), \( \partial \Omega \setminus \Gamma \)? The main result of the paper proves that if \((u_{0},B_{0})\) and \((u_{1},B_{1})\) are \(H^{r}\) regular divergence-free vector fields, where \(r\geq 3\) is an integer, with vanishing normal components on \((0,1)\times \{0,1\}\), and such that \(B_{0}\) and \(B_{1}\) further satisfy \(\int_{[0,1]^{2}}B_{0}^{1}dxdy= \int_{[0,1]^{2}}B_{1}^{1}dxdy=0\), there exists a solution \((u,B)\in C([0,T];H^{r}([0,1]^{2}))\) to the MHD equations which satisfies the different conditions hold. If the last hypothesis on the initial and terminal data is not satisfied, the same theorem holds but with a modified MHD equation for the magnetic field \(B\): \(\partial _{t}B+u\cdot \nabla B-B\cdot \nabla u=h(t)e_{x}\). The authors first extend the domain and the initial data \((u_{0},B_{0})\) to yield functions which are periodic in \(x\) on a larger domain and still satisfy the appropriate divergence-free and impermeability conditions. They then prove that for divergence-free vector fields on the square \([0,1]^{2}\) for which the mean of the first component vanishes, there is a divergence-free extension which vanishes on a large portion of \([-1,5]\times \lbrack 0,1]\). They finally observe that the magnetic field vanishes on \([0,1]^{2}\), whence the MHD on this domain reduces to the Euler equations, and they quote known control results for the Euler equations.