No label defined (Q3976488)

[For the entire collection Zbl 0734.00028.] The authors present a two-step technique based on the successive use of (analytical) perturbation expansions methods and (numerical) Bubnov- Galerkin approximations. They are considering problems of the form \(L(u,\varepsilon)=0\) in \(D\subset\mathbb{R}^ 2\), \(u=0\) on \(\partial D\), where \(L\) is a second order elliptic differential operator and \(\varepsilon\) is a parameter. In the first step coordinate functions are obtained by perturbation expansions of \(u\) about one or more specific values of the parameter \(\varepsilon\). In the second step some of these perturbation coordinate functions are used as trial functions in Bubnov-Galerkin approximations. The technique is applied to three model problems: a linear boundary value problem in a square, a nonlinear boundary value problem in an infinite domain and a nonlinear free boundary value problem. For the entire collection see [Zbl 0736.65059].