The indicatrix method in free boundary problems

Summary An approximate method for nonlinear problems with functional constraints is considered, in which the constraint in the whole domain is replaced by the constraint on a manifold of lower dimension. The stability criterion is introduced, and convergence theorems are proved for the onedimensional problem. Numerical results for the elastic-plastic torsion problem are given.     

Domain variations and moving boundary problems

In the past few decades maximal regularity theory has successfully been applied to moving boundary problems. The basic idea is to reduce the system with varying domains to one in a fixed domain. This is done by a transformation, the so-called Hanzawa transformation, and yields a typically nonlocal and nonlinear coupled system of (evolution) equations. Well-posedness results can then often be established as soon as it is proved that the relevant linearization is the generator of an analytic semigroup or admits maximal regularity. To implement this program, it is necessary to somehow parametrize to space of boundaries/domains (typically the space of compact hypersurfaces \(\Gamma \) in \({\mathbb {R}}^n\), in the Euclidean setting). This has traditionally been achieved by means of the already mentioned Hanzawa transformation. The approach, while successful, requires the introduction of a smooth manifold \(\Gamma _\infty \) close to the manifold \(\Gamma _0\) in which one cares to linearize. This prevents one to use coordinates in which \(\Gamma _0\) lies at their “center”. As a result formulæ tend to contain terms that would otherwise not be present were one able to linearize in a neighborhood emanating from \(\Gamma _0\) instead of from \(\Gamma _\infty \). In this paper it is made use of flows (curves of diffeomorphisms) to obtain a general form of the relevant linearization in combination with an alternative coordinatization of the manifold of hypersurfaces, which circumvents the need for the introduction of a “phantom” reference manifold \(\Gamma _\infty \) by, in its place, making use of a “phantom geometry” on \(\Gamma _0\). The upshot is a clear insight into the structure of the linearization, simplified calculations, and simpler formulæ for the resulting linear operators, which are useful in applications.     

Parameter estimation in moving boundary problems

We describe an approximation scheme which can be used to estimate unknown parameters in moving boundary problems. The model equations we consider are fairly general nonlinear diffusion/reaction equations of one spatial variable. Here we give conditions on the parameter sets and model equations under which we can prove that the estimates obtained using the approximations will converge to best-fit parameters for the original model equations. We conclude with a numerical example.     

On free boundary problems with moving contact points for the stationary two-dimensional Navier-Stokes equations

Solvability of the problem of slow drying of a plane capillary in the classical setting (i. e., with the adherence condition on a rigid wall) is established. The proof is based on a detailed study of the asymptotics of the solution near a point of contact of the free boundary with a moving wall, including estimates of the coefficients in well known asymptotic formulas. It is shown that the only value of the contact angle admitting a solution of the problem with finite energy dissipation equals π. Bibliography: 18 titles. Translated fromZapiski Nauchnykh Seminarov POMI, Vol. 213, 1994, pp. 179–205. Translated by E. V. Frolova.     

One version of the Linearized theory of nonstationary boundary-value problems with free boundary

We analyze the principle of linearization and linear boundary-value problems obtained by using this principle in the nonlinear theory of motion for a bounded volume of liquid with free surface subjected to the action of a nonstationary oscillating load. We formulate and study the problem of vibrocapillary equilibrium state, spectral problems in the theory of linear waves, and problems of stability of equilibrium states, including the problem of bifurcation of equilibrium states.     

Solvability of some two-dimensional quasistationary free boundary problems for the Navier-Stokes equations with moving contact points

The paper is concerned with some quasistationary two-dimensional free boundary problems of viscous flow with moving contact points and with contact angle equal to π. A typical example of such a flow is filling a capillary tube in the presence of surface tension. The proof of the solvability of these problems is based on the analysis (made by the author and V. V. Pukhnachëv about 10 years ago) of the asymptotic formulas for the solutions of the Navier-Stokes equations in a neigborhood of contact points. Bibliography: 10 titles.     

A free boundary problem with moving source

A free boundary problem with a moving source is considered. The existence of a critical velocity, above which a fault line occurs, is known experimentally in welding. One possible explanation has been the formation of a teardrop-shaped weld pool with a sharp vertex at the tail end. It is proved that such a singularity cannot occur within a reasonable two-dimensional model. The model includes physical effects due to discontinuous diffusivities, convection terms, and deviations in the interfacial temperature due to terms such as curvature and linear dynamics.     

Two related free boundary problems

Two related free boundary problems are solved: the first isthe viscous film coating of wedges of arbitrary angle; the secondis the rectangular dam problem with evaporation (or fluid removal)from the free surface. Both problems are of practical interestand explicit solutions are given. The two examples treated aregeneralizations of problems solved using Polubarinova-Kochina's(1962) analytic differential equation method and conformal mappingsinvolving elliptic modular functions to an intermediate plane.Here conformal mappings involving Legendre functions are usedto generalize these results.     

On free boundary problems in two dimensions

Let L be a linear elliptic operator in two dimensions with analytic coefficients and of second order, andu(x, y) a solution of Lu=0 in a simply connected domain ω with rectifiable boundary Γ. Suppose ψ(x, y) analytic on ω∪Γ and L ψ≠0 there.H is shown that ifu and ψ coincide with first derivatives on an open portion Γ₀ of Γ, then Γ₀ permits the representation λ=x (θ),y=y (θ) withx(θ),y(θ)analytic functions of a real parameter θ.     

Certain problems of the theory of difference boundary value problems

This is an author's abstract of a dissertation for the degree of doctor of physicomathematical sciences. The dissertation was defended on December 2,1969,at a session of the Scientific Council of the Institute of Applied Mathematics, Academy of Sciences of the USSR. The official opponents were: Doctor of Physicomathematical Sciences Prof. K. I. Babenko, Doctor of Physicomathematical Sciences Prof. N. S. Bakhvalov, and Doctor of Physicomathematical Sciences E. A. Volkov.Translated from Matematicheskie Zametki, Vol. 7, No. 5, pp. 655–663, May, 1970.     

Boundary value problems in the theory of analytic functions with loaded free terms and additionally given boundary moments

Doklady Mathematics -     

Efficient treatment of stationary free boundary problems

In the present paper we consider the efficient treatment of free boundary problems by shape optimization. We reformulate the free boundary problem as shape optimization problem. A second order shape calculus enables us to analyze the shape problem under consideration and to prove convergence of a Ritz–Galerkin approximation of the shape. We show that Newton's method requires only access to the underlying state function on the boundary of the domain. We compute these data by boundary integral equations which are numerically solved by a fast wavelet Galerkin scheme. Numerical results prove that we succeeded in finding a fast and robust algorithm for solving the considered class of problems.