A quasi-periodic Poincaré's theorem

We study the persistence of invariant tori on resonant surfaces of a nearly integrable Hamiltonian system under the usual Kolmogorov non-degenerate condition. By introducing a quasi-linear iterative scheme to deal with small divisors, we generalize the Poincaré theorem on the maximal resonance case (i.e., the periodic case) to the general resonance case (i.e., the quasi-periodic case) by showing the persistence of majority of invariant tori associated to non-degenerate relative equilibria on any resonant surface.The first author was partially supported by NSFC grant 19971042, the National 973 Project of China: Nonlinearity, and the outstanding young's project of the Ministry of Education of China.The second author was partially supported by NSF grant DMS9803581.Mathematics Subject Classification (2000): Primary 58F05, 58F27, 58F30     

Analytical reduction for a concentration dependent diffusion problem

Previous analytical simplifications of Boltzmann's similarity solution are unified and generalized by examining symmetry and transformation properties of the equation. It is shown how to reduce boundary value problems to initial value problems for certain diffusivities. A series solution is derived for this case, and shown to be accurate for a broad range of parameter values.

---

Résumé Les simplifications analytiques connues de la solution similaire de Boltzmann sont unifiées et généralisées par l'étude des propriétés de symétrie et des transformations de l'équation de diffusion. On montre comment, pour certaines diffusivités, un problème de conditions aux limites se réduit à un problème de valeurs initiales. Une solution en séries obtenue pour ce cas est très précise pour un domaine étendu des paramètres.

     

Symbolic-computation study of integrable properties for the (2 + 1)-dimensional Gardner equation with the two-singular manifold method

The singular manifold method from the Painlevé analysiscan be used to investigate many important integrable propertiesfor the non-linear partial differential equations. In this paper,the two-singular manifold method is applied to the (2 + 1)-dimensionalGardner equation with two Painlevé expansion branchesto determine the Hirota bilinear form, Bäcklund transformation,Lax pairs and Darboux transformation. Based on the obtainedLax pairs, the binary Darboux transformation is constructedand the N x N Grammian solution is also derived by performingthe iterative algorithm N times with symbolic computation.     

Nonintegrability of a Fifth-Order Equation with Integrable Two-Body Dynamics

We consider a fifth-order partial differential equation (PDE) that is a generalization of the integrable Camassa–Holm equation. This fifth-order PDE has exact solutions in terms of an arbitrary number of superposed pulsons with a geodesic Hamiltonian dynamics that is known to be integrable in the two-body case N==2. Numerical simulations show that the pulsons are stable, dominate the initial value problem, and scatter elastically. These characteristics are reminiscent of solitons in integrable systems. But after demonstrating the nonexistence of a suitable Lagrangian or bi-Hamiltonian structure and obtaining negative results from Painlevé analysis and the Wahlquist–Estabrook method, we assert that this fifth-order PDE is not integrable.     

A criterion for non-integrability of Hamiltonian systems with nonhomogeneous potentials

A necessary condition for integrability is presented for two degrees of freedom Hamiltonian systems with nonhomogeneous potentials. This is based on a theorem of Ziglin and gives also an efficient criterion for non-integrability. As an application, it is used to single out integrable cases among the Hénon-Heiles Hamiltonians.

---

Résumé On démontre une condition nécessaire pour l'intégrabilité de systèmes hamiltoniens à deux degrés de liberté qui possèdent des potentiels inhomogènes. La démonstration repose sur un théorème de Ziglin et l'on obtient aussi un critère de non-integrabilité. Comme application, on s'en sert pour trouver des cas intégrables dans les hamiltoniens de Hénon-Heiles.

---

---

Dedicated to Prof. Yoshikazu Hirasawa on his 60th birthday

---

This research was partially supported by Grant-in-Aid for Scientific Research (No. 60740069), Ministry of Education. Japan.

---

Staying at Forschungsinstitut für Mathematik, ETH-Zentrum, Zürich until the end of August, 1987.     

Necessary Covariance Conditions for a One-Field Lax Pair

We study the covariance with respect to Darboux transformations of polynomial differential and difference operators with coefficients given by functions of one basic field. In the scalar (Abelian) case, the functional dependence is established by equating the Frechet differential (the first term of the Taylor series on the prolonged space) to the Darboux transform; a Lax pair for the Boussinesq equation is considered. For a pair of generalized Zakharov-Shabat problems (with differential and shift operators) with operator coefficients, we construct a set of integrable nonlinear equations together with explicit dressing formulas. Non-Abelian special functions are fixed as the fields of the covariant pairs. We introduce a difference Lax pair, a combined gauge-Darboux transformation, and solutions of the Nahm equations.__________Translated from Teoreticheskaya i Matematicheskaya Fizika, Vol. 144, No. 1, pp. 122–132, July, 2005.     

Stability of flow of a generalized Newtonian fluid down an inclined plane

Summary The analogue of Orr-Sommerfeld equation is derived for a generalized Newtonian fluid. Based on this equation, the stability of such fluid flowing down an inclined plane under gravity is studied. The critical Reynolds number is given as a function of dimensionless steady flow velocityU(y) and the slope of the plane, and is computed for several fluids.

---

Résumé On a étudié le problème de stabilité de l'écoulement d'un fluide Newtonien généralisé sur un plan inclinè.

     

Periodic Liénard-type delay equations with state-dependent impulses

Conditions are obtained for Liénard-type equations with delay and state-dependent impulses to admit an absolutely continuous periodic solution with first derivative of bounded variation (and consequently with Lebesgue integrable second derivative). The results are applied to Josephson's equation and the nonconservative forced pendulum equation.     

Application of average dynamic programming to inventory systems

We show the existence ofaverage cost (AC-) optimal policy for an inventory system withuncountable state space; in fact, the AC-optimal cost and an AC-optimal stationary policy areexplicitly computed. In order to do this, we use a variant of thevanishing discount factor approach, which have been intensively studied in recent years but the available results not cover the inventory problem we are interested in.The work of the first author (OVA) was partially supported by Fondo del Sistema de Investigación del Mar de Cortéz under grant SIMAC/94/CT-005. The work of the second author (RMdO) was partially supported by Consejo Nacional de Ciencia y Tecnologia (CONACyT) under grant 0635P-E9506.     

Shape sensitivity analysis via a penalization method

Summary The object of this paper is the development of a penalization technique to compute the shape derivative of cost functionals where the state is the solution of a non-linear equation and/or a linear variational inequality. This type of problem is frequently encountered in Shape Sensitivity Analysis.

---

Résumé Cet article présente le calcul des dérivées de forme de fonctionnelles définies sur un domaine géométrique par une méthode de pénalisation. On suppose que l'état est la solution d'une équation non-linéaire ou d'une inéquation linéaire. Ce type de problème est fréquemment rencontré en analyse de sensitivité des formes.

---

---

This research was supported in part by the National Sciences and Engineering Council of Canada Operating Grant A-8730 and a FCAR Grant from the « Ministère de l'Education du Québec ».     

Invariant Simultaneous Solutions of Evolution Equations of Integrable Hierarchies

We construct classical point symmetry groups for joint pairs of evolution equations (systems of equations) of integrable hierarchies related to the auxiliary equation of the method of the inverse problem of second order. For the two cases: the hierarchy of Korteweg--de Vries (KdV) equations and of the systems of Kaup equations, we construct simultaneous solutions invariant with respect to the symmetry group. The problem of the construction of these solutions can be reduced, respectively, to the first and second Painlevé equations depending on a parameter. The Painlevé equations are supplemented by the linear evolution equations defining the deformation of the solution of the corresponding Painlevé equation.     

The multi-soliton solutions of the CH-γ equation

This paper solves the integrable CH-γ equation for analytical multiple soliton solutions with the Darboux transformation method. Some properties of the soliton solutions are different from the CH equation. This work was partially supported by the National Natural Science Foundation of China (Grant No. 10401022) and the Research Grants Council of Hong Kong     

A new kind of nonlocal symmetry for the μ‐Camassa‐Holm equation with linear dispersion

The μ‐Camassa‐Holm equation with linear dispersion is a completely integrable model. In this paper, it is shown that this equation has quadratic pseudo‐potentials that allow us to construct pseudo‐potential–type nonlocal symmetries. As an application, we obtain its recursion operator by using this kind of nonlocal symmetry, and we construct a Darboux transformation for the μ‐Camassa‐Holm equation.     

One variant of a (2 + 1)-dimensional Volterra system and its (1 + 1)-dimensional reduction

A new system is generated from a multi-linear form of a (2+1)-dimensional Volterra system. Though the system is only partially integrable and needs additional conditions to possess two-soliton solutions, its (1+1)-dimensional reduction gives an integrable equation which has been studied via reduction skills. Here, we give this (1+1)-dimensional reduction a simple bilinear form, from which a Bäcklund transformation is derived and the corresponding nonlinear superposition formula is built.     

Complete integrability and the Miura transformation of a coupled KdV equation

In this letter, a Painlevé integrable coupled KdV equation is proved to be also Lax integrable by a prolongation technique. The Miura transformation and the corresponding coupled modified KdV equation associated with this equation are derived.     

Bianchi Permutability for the Anti‐Self‐Dual Yang‐Mills Equations

The anti‐self‐dual Yang‐Mills equations are known to have reductions to many integrable differential equations. A general Bäcklund transformation (BT) for the anti‐self‐dual Yang‐Mills (ASDYM) equations generated by a Darboux matrix with an affine dependence on the spectral parameter is obtained, together with its Bianchi permutability equation. We give examples in which we obtain BTs of symmetry reductions of the ASDYM equations by reducing this ASDYM BT. Some discrete integrable systems are obtained directly from reductions of the ASDYM Bianchi system.     

On the mixed boundary-value problem for harmonic functions in plane domains

This paper considers the problem of the determination of a harmonic function in a simply connected plane domain when the values of the function are known on some arcs of the boundary and the values of the normal derivative are known on the remaining boundary. We first present the solution in theoretical form and then show how to compute the solution with the aid of rapidly convergent series. The coefficients of these series are Fourier coefficients of certain functions and can be estimated by using the Fast Fourier Transform. The examples considered in the last part of the paper emphasize the advantages of the method presented in this paper for solving the mixed boundary-value problem as compared to other methods used for this purpose.

---

Résumé On considère le problème suivant: trouver une fonction harmonique dans un domaine simplement connexeDR ² dont la frontièreS est assez régulière, en connaissant les valeurs de la fonction sur une partieS ₁ de la frontière et les valeurs de la dérivée normale sur la complémentaire de366-1. La première partie présente des résultats théoriques liés à ce probléme. La deuxième partie s'appuie sur une représentation de la solution à l'aide de certaines séries de fonctions rapidement convergentes. Les coefficients des séries utilisées sont même les coefficients de Fourier de certaines fonctions, et leur calcul peut être effectué en utilisant la transformation de Fourier rapide (Fast Fourier Transform). Les exemples considérés dans la dernière partie de l'article mettent en évidence les avantages de la méthode présentée par rapport à d'autres méthodes employées pour la résolution du problème mixte.

     

Isothermic Surfaces in Spaces of Arbitrary Dimension: Integrability, Discretization, and Bäcklund Transformations—A Discrete Calapso Equation

A vector analog of the classical Calapso equation governing isothermic surfaces in R ^{n +2} is introduced. It is shown that this vector Calapso system admits a nonlocal) scalar Lax pair based on the classical Moutard equation. The analog of Darboux's Bäcklund transformation for isothermic surfaces in R³ is derived in a systematic manner and shown that it may be formulated in terms of the classical Moutard transformation acting on the scalar Lax pair. A permutability theorem for isothermic surfaces is set down that manifests itself in an explicit superposition principle for the vector Calapso system. This superposition principle in vectorial form is shown to constitute an integrable discretization of the vector Calapso system and, therefore, defines discrete isothermic surfaces in R ^{n +2}. The discrete Calapso equation is related to the discrete Korteweg–de Vries equation and discrete holomorphic functions. A matrix Lax pair based on Clifford algebras and a scalar Lax pair are derived for the discrete Calapso equation. A discrete Moutard-type transformation for the discrete Calapso equation is obtained, and it is shown that the discrete Calapso equation may be specialized to an integrable discrete version of the O( n +2) nonlinear σ-model.     

Multiple complex soliton solutions for integrable negative-order KdV and integrable negative-order modified KdV equations

In this work we show that the integrable negative-order Korteweg–de Vries (nKdV) and the integrable negative-order modified Korteweg–de Vries (nMKdV) equation admit multiple complex soliton solutions. To achieve this goal, we introduce two complex forms of the simplified Hirota’s method, the first works effectively for the nKdV equation, and the other form is nicely applicable for the nMKdV equation. We believe that the newly proposed complex forms and the obtained findings will shed light on complex solitons of other integrable equations.     

Periodic fluid transients in cylindrical ducts with transverse magnetic fields

Summary This paper treats periodic fluid transients in liquid metals contained in constant-area circular ducts with uniform, transverse, applied magnetic fields. The magnetic Reynolds and Mach numbers are assumed to be small, and the Reynolds number is assumed to be large. By the use of a complex coordinate transformation, closed-form solutions are obtained for the perfectly conducting duct. By the same technique, solutions in series of Mathieu functions are obtained for the fully insultating duct. A method for solving the dispersion relation for each case is outlined, and sample numerical results are presented.

---

Résumé Cet article traite des ondes de compression périodiques dans un métal liquide contenu dans une conduite circulaire avec une section droite qui est constante et avec un champ magnétique transversal et uniforme que l'on applique. Ici on suppose que le nombre magnétique de Reynolds et le nombre de Mach sont petits et que le nombre de Reynolds est grand. A l'aide d'une transformation complexe des corrdonnées on obtient des solutions éxactes pour une conduite qui est parfaitement conductrice. Par le même technique on trouve les solutions pour la conduite parfaitement isolante dans une série de fonctions de Mathieu. Une méthode pour la solution de l'équation de dispersion est esquissée et des exemples de résultats numériques sont présentés.