finite element methods;
maximum principles;
discrete maximum principles;
quasi-linear elliptic equations;
D O I:
10.1137/110833737
中图分类号:
O29 [应用数学];
学科分类号:
070104 ;
摘要:
This paper derives some discrete maximum principles for P1-conforming finite element approximations for quasi-linear second order elliptic equations. The results are extensions of the classical maximum principles in the theory of partial differential equations to finite element methods. The mathematical tools are based on the variational approach that was commonly used in the classical partial differential equation theory. The discrete maximum principles are established by assuming a property on the discrete variational form that is of global nature. In particular, the assumption on the variational form is verified when the finite element partition satisfies some angle conditions. For the general quasi-linear elliptic equation, these angle conditions indicate that each triangle or tetrahedron needs to be O(h(alpha))-acute in the sense that each angle alpha(ij) (for the triangle) or interior dihedral angle a(ij) (for the tetrahedron) must satisfy alpha(ij) <= pi/2 - gamma h(alpha) for some alpha >= 0 and gamma > 0. For the Poisson problem where the differential operator is given by the Laplacian, the angle requirement is the same as the existing ones: either all the triangles are nonobtuse or each interior edge is nonnegative. It should be pointed out that the analytical tools used in this paper are based on the powerful De Giorgi iterative method that has played important roles in the theory of partial differential equations. The mathematical analysis itself is of independent interest in the finite element analysis.
机构:
South Asian Univ, Fac Math & Comp Sci, Dept Appl Math, New Delhi 110021, IndiaSouth Asian Univ, Fac Math & Comp Sci, Dept Appl Math, New Delhi 110021, India
Mohanty, Ranjan K.
Setia, Nikita
论文数: 0引用数: 0
h-index: 0
机构:
Univ Delhi, Fac Math Sci, Dept Math, Delhi 110007, IndiaSouth Asian Univ, Fac Math & Comp Sci, Dept Appl Math, New Delhi 110021, India