Chain recurrence,semiflows, and gradients

This paper is a study of chain recurrence and attractors for maps and semiflows on arbitrary metric spaces. The main results are as follows. (i) C. Conley's characterization of chain recurrence in terms of attractors holds for maps and semiflows on any metric space. (ii) An alternative definition of chain recurrence for semiflows is given and is shown to be equivalent to the usual definition. The alternative definition uses chains formed of orbit segments whose lengths are at least 1, while in the usual definition these lengths are required to be arbitrarily long. (iii) The chain recurrent set of a continuous semiflow is the same as the chain recurrent set of its time-one map. (iv) Conditions on a real-valued function are given that ensure that the semiflow generated by its gradient has only equilibria in its chain recurrent set. An example is given (onR ³) showing that a gradient flow may have nonequilibrium chain recurrent points if these conditions are violated.     

Stable Periodic Motion of a System Using Echo for Position Control

We study a system of equations which is based on Newton's law and models automatic position control by echo. Rewritten as a delay differential equation with state-dependent delay the model defines a semiflow on a submanifold in the space of continuously differentiable maps [-h, 0] ². All time-t maps and the restriction of the semiflow given by t > h are continuously differentiable. For simple nonlinearities and suitable parameters we find a set of initial data to which the solution curves return, after an excursion into the ambient manifold. The associated return map is semiconjugate to an interval map. Estimates of derivatives yield a unique, attracting fixed point of the latter, which can be lifted to the return map. The proof that the resulting periodic orbit is stable and exponentially attracting with asymptotic phase involves a Poincaré return map and a discussion of derivatives of its iterates.     

Energy Flow in Formally Gradient Partial Differential Equations on Unbounded Domains

As an example of an extended, formally gradient dynamical system, we consider the damped hyperbolic equation u _tt+u _t=u+F(x, u) in R ^N , where F is a locally Lipschitz nonlinearity. Using local energy estimates, we study the semiflow defined by this equation in the uniformly local energy space H¹ _ul(R ^N )×L² _ul(R ^N ). If N2, we show in particular that there exist no periodic orbits, except for equilibria, and we give a lower bound on the time needed for a bounded trajectory to return in a small neighborhood of the initial point. We also prove that any nonequilibrium point has a neighborhood which is never visited on average by the trajectories of the system, and we conclude that any bounded trajectory converges on average to the set of equilibria. Some counter-examples are constructed, which show that these results cannot be extended to higher space dimensions.     

Spinodal Decomposition for the¶Cahn-Hilliard Equation in Higher Dimensions:¶Nonlinear Dynamics

This paper addresses the phenomenon of spinodal decomposition for the Cahn-Hilliard equation

Uniqueness of positive solutions of Δu−u+up=0 in Rn

We establish the uniqueness of the positive, radially symmetric solution to the differential equation u–u+u^p=0 (with p>1) in a bounded or unbounded annular region in R ⁿ for all n1, with the Neumann boundary condition on the inner ball and the Dirichlet boundary condition on the outer ball (to be interpreted as decaying to zero in the case of an unbounded region). The regions we are interested in include, in particular, the cases of a ball, the exterior of a ball, and the whole space. For p=3 and n=3, this a well-known result of Coffman, which was later extended by McLeod & Serrin to general n and all values of p below a certain bound depending on n. Our result shows that such a bound on p is not needed. The basic approach used in this work is that of Coffman, but several of the principal steps in the proof are carried out with the help of Sturm's oscillation theory for linear second-order differential equations. Elementary topological arguments are widely used in the study.     

Ergodic Properties of Weak Asymptotic Pseudotrajectories for Semiflows

It is known that various deterministic and stochastic processes such as asymptotically autonomous differential equations or stochastic approximation processes can be analyzed by relating them to an appropriately chosen semiflow. Here, we introduce the notion of a stochastic process X being a weak asymptotic pseudotrajectory for a semiflow and are interested in the limiting behavior of the empirical measures of X. The main results are as follows: (1) the weak* limit points of the empirical measures for X axe almost surely -invariant measures; (2) given any semiflow , there exists a weak asymptotic pseudotrajectory X of such that the set of weak* limit points of its empirical measures almost surely equal the set of all ergodic measures for ; and (3) if X is an asymptotic pseudotrajectory for a semiflow , then conditions on that ensure convergence of the empirical measures are derived.     

Removable Singularities for Fully Nonlinear Elliptic Equations

We obtain a removability result for the fully nonlinear uniformly elliptic equations F(D ² u)+f(u)=0. The main theorem states that every solution to the equation in a punctured ball (without any restrictions on the behaviour near the centre of the ball) is extendable to the solution in the entire ball provided the function f satisfies certain sharp conditions depending on F. Previously such results were known for linear and quasilinear operators F. In comparison with the semi- or quasilinear theory the techniques for the fully nonlinear equations are new and based on the use of the viscosity notion of generalised solution rather than the distributional or the weak solutions. Accepted May 3, 2000?Published online November 16, 2000     

Finite dimensional aspects of semilinear parabolic equations

We prove that the semiflow generated by _t ,u– _xx ² u=f(x, u), withu(0, t)= u(1,t)=0, onL ²([0, 1]) isC -conjugate to a product of finite dimensional flows.     

Point Singularities of 3D Stationary Navier–Stokes Flows

This article characterizes the singularities of very weak solutions of 3D stationary Navier–Stokes equations in a punctured ball which are sufficiently small in weak L ³.     

Quasi-Periodic Stability of Subfamilies of an Unfolded Skew Hopf Bifurcation

In the skew Hopf bifurcation a quasi-periodic attractor with nontrivial normal linear dynamics loses hyperbolicity. The simplest setting concerns rotationally symmetric diffeomorphisms of S ¹×R ². Their dynamics involve periodicity, quasi-periodicity and chaos, including mixed spectrum. The present paper deals with the persistence under symmetry-breaking of quasi-periodic invariant circles in this bifurcation. It turns out that, when adding sufficiently many unfolding parameters, the invariant circle persists for a large Hausdorff measure subset of a submanifold in parameter space. Accepted: September 3, 1999     

Regularity of invariant manifolds

Under mild conditions it is proved that an invariant submanifold ofX 0<1 for the equationdx/dt+Ax=f(x), A sectorial,fC'(X ,X),0<1, is a submanifold ofX ¹ as well. In addition, conditions are given for the semiflow of the equation to extend fromX toX and a new inertial manifold theorem is proved for the scalar reaction diffusion equation.     

Energy Minimising Properties of the Radial Cavitation Solution in Incompressible Nonlinear Elasticity

Consider an incompressible, hyperelastic material occupying the unit ball B⊂ℝ ⁿ in its reference state. Suppose that the deformation u:B→ℝ ⁿ is specified on the boundary by

The Asymptotic Properties of the Solution to the Stokes Problem in Domains That Are Layer-Like at Infinity

Asymptotic formulae are derived for solutions to the Stokes problem in domains which, outside a ball, coincide with the three-dimensional layer \Bbb R² ×(0,1) {\Bbb R}^2 \times (0,1) . The properties of detached asymptotic terms differ in the transversal and longitudinal directions. In order to justify the asymptotic expansions the procedure of dimension reduction is employed together with estimates for miscellaneous weighted norms of the solutions.     

Existence of Compressible Current-Vortex Sheets: Variable Coefficients Linear Analysis

We study the initial-boundary value problem resulting from the linearization of the equations of ideal compressible magnetohydrodynamics and the Rankine-Hugoniot relations about an unsteady piecewise smooth solution. This solution is supposed to be a classical solution of the system of magnetohydrodynamics on either side of a surface of tangential discontinuity (current-vortex sheet). Under some assumptions on the unperturbed flow, we prove an energy a priori estimate for the linearized problem. Since the tangential discontinuity is characteristic, the functional setting is provided by the anisotropic weighted Sobolev space W₂^1,σ. Despite the fact that the constant coefficients linearized problem does not meet the uniform Kreiss-Lopatinskii condition, the estimate we obtain is without loss of smoothness even for the variable coefficients problem and nonplanar current-vortex sheets. The result of this paper is a necessary step in proving the local-in-time existence of current-vortex sheet solutions of the nonlinear equations of magnetohydrodynamics.     

Elliptic Equations in Divergence Form with Partially BMO Coefficients

The solvability in Sobolev spaces is proved for divergence form second order elliptic equations in the whole space, a half space, and a bounded Lipschitz domain. For equations in the whole space or a half space, the leading coefficients a ^ij are assumed to be only measurable in one direction and have locally small BMO semi-norms in the other directions. For equations in a bounded domain, additionally we assume that a ^ij have small BMO semi-norms in a neighborhood of the boundary of the domain. We give a unified approach of both the Dirichlet boundary problem and the conormal derivative problem. We also investigate elliptic equations in Sobolev spaces with mixed norms under the same assumptions on the coefficients.     

Stress state of an elastic ball in a spherically symmetric gravitational field

We consider a spherically symmetric static problem of general relativity whose solution was obtained in 1916 by Schwarzschild for a metric form of a special type. This solution determines the metric coefficients of the exterior and interior Riemannian spaces generated by a gravitating solid ball of constant density and includes the so-called gravitational radius r _g. For a ball of outer radius R=r _g, the metric coefficients are singular, and hence the radius r _g is traditionally assumed to be the radius of the event horizon of an object called a black hole. The solution of the interior problem obtained for an incompressible ideal fluid shows that the pressure at the ball center increases without bound for R=9/8r _g, which is traditionally used for the physical justification of the existence of black holes. The discussion of Schwarzschild’s traditional solution carried out in this paper shows that it should be generalized with respect to both the geometry of the Riemannian space and the elastic medium model. In this connection, we consider the general metric form of a spherically symmetric Riemannian space and prove that the solution of the corresponding static problem exists for a broad class of metric forms. A special metric form based on the assumption that the gravitation generating the Riemannian space inside a fluid ball or an elastic ball does not change the ball mass is singled out from this class. The solution obtained for the special metric form is singular with respect to neither the metric coefficients nor the pressure in the fluid ball and the stresses in the elastic ball. The obtained solution is compared with Schwarzschild’s traditional solution.     

The Space-Time Finite Element Method for Parabolic Problems

ＩｎｔｒｏｄｕｃｔｉｏｎＴｈｅｅｑｕａｔｉｏｎｓｗｅｃｏｎｓｉｄｅｒｅｄａｒｅａｓｆｏｌｌｏｗｓｕｔ-Δｕ =ｆ(ｕ) ,　　Ω× [0 ,Ｔ] ,ｕ｜ Ω =0 ,　　　　　 Ω× [0 ,Ｔ] ,ｕ( · ,0 ) =ｕ0 ,Ω ,( 1 )ｗｈｅｒｅΩ ∈Ｒ2 ,ｔｈｅｆｕｎｃｔｉｏｎｆ(ｕ)ｓａｔｉｓｆｉｅｓ｜ｆ(ｕ)｜≤ｃ｜ｕ｜ ,　　 ｕ∈Ｃ(Ω) . ( 2 )Ａｎｄｆ(ｕ)ｉｓＬｉｐｓｃｈｉｔｚｃｏｎｔｉｎｕｏｕｓ,ｉ.ｅ.ｉｔｓａｔｉｓｆｉｅｓ｜ｆ(ｕ) -ｆ(ｖ) ｜≤Ｌ｜ｕ-ｖ｜ ,　　 ｕ ,ｖ∈Ｃ(Ω) ,( 3 )ｗｈｅｒｅＬｉｓＬｉｐｓｃｈｉｔｚｃｏｎｓｔａ…     

On Large-Time Behavior of Solutions to the Compressible Navier-Stokes Equations in the Half Space in R 3

 The Navier-Stokes equation for compressible viscous fluid is considered on the half space in R ³ under the zero-Dirichlet boundary condition for the momentum with initial data near an arbitrarily given equilibrium of positive constant density and zero momentum. Time decay properties in L ² norms for solutions of the linearized problem are investigated to obtain the rate of convergence in L ² norms of solutions to the equilibrium when initial data are sufficiently close to the equilibrium in . Some lower bounds are derived for solutions to the linearized problem, one of which indicates a nonlinear phenomenon not appearing in the case of the Cauchy problem on the whole space. (Accepted May 8, 2002) Published online October 18, 2002 Communicated by T.-P. LIU     

On the orders of the best approximations of integrals of functions by integrals of rank σ

We study the values e _σ(f) of the best approximation of integrals of functions from the spaces L _p (A, dμ) by integrals of rank σ. We determine the orders of the least upper bounds of these values as σ → ∞ in the case where the function ƒ is the product of two nonnegative functions one of which is fixed and the other varies on the unit ball U _p (A) of the space L _p (A, dμ). We consider applications of the obtained results to approximation problems in the spaces S _p ^ϕ. __________ Translated from Neliniini Kolyvannya, Vol. 10, No. 4, pp. 528–559, October–December, 2007.     

Flow structure of knuckling effect in footballs

The flight trajectory of a non-spinning or slow-spinning soccer ball might fluctuate in unpredictable ways, as for example, in the many free kicks of C. Ronaldo. Such anomalous horizontal shaking or rapid falling is termed the ‘knuckling effect’. However, the aerodynamic properties and boundary-layer dynamics affecting a ball during the knuckling effect are not well understood. In this study, we analyse the characteristics of the vortex structure of a soccer ball subject to the knuckling effect (knuckleball), using high-speed video images and smoke-generating agents. Two high-speed video cameras were set at one side and in front of the ball trajectory between the ball position and the goal; further, photographs were taken at 1000 fps and a resolution of 1024×512 pixels. Although in a previous study (Taneda, 1978), shedding of horseshoe vortices was observed for smooth spheres in the Reynolds number (Re) range of 3.8×10⁵<Re<10⁶, in the case of the soccer ball, the vortex structure, which consisted of distorted loop vortices, appeared in the wake behind the ball in the supercritical Re number region. Moreover, after the knuckleballs were airborne, large-scale undulations were observed in the vortex trail visualised with a smoke technique. On the other hand, aerodynamic forces acting on the ball were estimated from the data of the ball’s flight trajectory, and a statistically high correlation (r=0.94, p<0.01) between the fluctuation frequency of the lift and side forces and the undulation frequency of the vortex trail was shown to exist. This fact suggests that the phenomenon of large-scale undulations of the vortex trail is closely related to the cause of the unsteady aerodynamic forces acting on the knuckle ball.