Introduction
The efficient numerical treatment of boundary and transmission conditions is constantly an interesting subject that can
have many applications, for instance in domain decomposition methods. Several approaches are possible: These conditions
can be treated as essential boundary conditions. This means that they are explicitly included in the definition
of the function spaces, that corresponds more or less to the classical approach. Another way consists in approximating
boundary conditions by introducing Lagrange multipliers. This can be found in [2] for the Dirichlet problem.
Another approach which is of interest is related to the Nitsche Method [12], transferred to continuity conditions by
Stenberg. Several years ago, J. Nitsche introduced a method to impose weakly essential boundary conditions in the
scalar Laplace operator. This formulation has several advantages: It is well adapted to conforming finite element,
it is an efficient way to reuse available codes, built on conforming finite element methods. In addition, the Nitsche
formulation leads to a symmetric, definite, positive discrete formulation, in agreement with symmetry and ellipticity
Introduction
The efficient numerical treatment of boundary and transmission conditions is constantly an interesting subject that can
have many applications, for instance in domain decomposition methods. Several approaches are possible: These conditions
can be treated as essential boundary conditions. This means that they are explicitly included in the definition
of the function spaces, that corresponds more or less to the classical approach. Another way consists in approximating
boundary conditions by introducing Lagrange multipliers. This can be found in [2] for the Dirichlet problem.
Another approach which is of interest is related to the Nitsche Method [12], transferred to continuity conditions by
Stenberg. Several years ago, J. Nitsche introduced a method to impose weakly essential boundary conditions in the
scalar Laplace operator. This formulation has several advantages: It is well adapted to conforming finite element,
it is an efficient way to reuse available codes, built on conforming finite element methods. In addition, the Nitsche
formulation leads to a symmetric, definite, positive discrete formulation, in agreement with symmetry and ellipticity
การแปล กรุณารอสักครู่..
