N-Homoclinic bifurcations for homoclinic orbits changing their twisting

We study bifurcations, calledN-homoclinic bifurcations, which produce homoclinic orbits roundingN times (N2) in some tubular neighborhood of original homoclinic orbit. A family of vector fields undergoes such a bifurcation when it is a perturbation of a vector field with a homoclinic orbit.N-Homoclinic bifurcations are divided into two cases; one is that the linearization at the equilibrium has only real principal eigenvalues, and the other is that it has complex principal eigenvalues. We treat the former case, espcially that linearization has only one unstable eigenvalue. As main tools we use a topological method, namely, Conley index theory, which enables us to treat more degenerate cases than those studied by analytical methods.     

Central configurations of the N-body problem via equivariant Morse theory

In this paper we use the equivariant Morse theory to give an estimate of the minimal number of central configurations in the N-body problem in ³. In the case of equal masses we prove that the planar central configurations are saddle points for the potential energy. From this we deduce the presence of non-planar central configurations, for every N 4.The principal difficulty in applying Morse theory is that the potential function is defined on a manifold on which the group O(3) does not act freely. To avoid this problem the equivariant cohomology functor is applied in order to obtain the Morse inequalities.     

Homoclinic Saddle-Node Bifurcations in Singularly Perturbed Systems

In this paper we study the creation of homoclinic orbits by saddle-node bifurcations. Inspired on similar phenomena appearing in the analysis of so-called localized structures in modulation or amplitude equations, we consider a family of nearly integrable, singularly perturbed three dimensional vector fields with two bifurcation parameters a and b. The O() perturbation destroys a manifold consisting of a family of integrable homoclinic orbits: it breaks open into two manifolds, W ^s() and W ^u(), the stable and unstable manifolds of a slow manifold . Homoclinic orbits to correspond to intersections W ^s()W ^u(); W ^s()W ^u()= for a<a*, a pair of 1-pulse homoclinic orbits emerges as first intersection of W ^s() and W ^u() as a>a*. The bifurcation at a=a* is followed by a sequence of nearby, O( ²(log)²) close, homoclinic saddle-node bifurcations at which pairs of N-pulse homoclinic orbits are created (these orbits make N circuits through the fast field). The second parameter b distinguishes between two significantly different cases: in the cooperating (respectively counteracting) case the averaged effect of the fast field is in the same (respectively opposite) direction as the slow flow on . The structure of W ^s()W ^u() becomes highly complicated in the counteracting case: we show the existence of many new types of sometimes exponentially close homoclinic saddle-node bifurcations. The analysis in this paper is mainly of a geometrical nature.     

The self-preservation of homogeneous shear flow turbulence

An analysis of the equations governing homogeneous shear flow shows the possibility of solutions which are self-preserving at all scales of motion, and that these solutions are dependent on the initial conditions. The appropriate velocity scale is the one obtained from the turbulence kinetic energy, q ²/2, while the length scale is the Taylor microscale, . Two cases of self-preserving flow are identified: one corresponding to constant mean shear, the other to a mean shear which is inversely proportional to time. For the first case (the only one considered in detail) the principal results of the postulated similarity are that is constant, while q ² varies exponentially with time. The ratio of the turbulence energy production rate to its dissipation rate remains constant. It is also shown that the energy spectra scale over all wavenumbers with q ² and , and that they have shapes determined by the initial conditions. The experimental evidence is generally consistent with the theory.     

Comparison of stability and control characteristics of two twin-lift helicopter configurations

The stability and control characteristics of two twin-lift helicopter configurations are analyzed in this paper. In order to address the issue of configuration selection from a handling qualities viewpoint, their open-and closed-loop characteristics are compared. The two twin-lift configurations considered are the twin-lift with spreader bar and twin-lift without spreader bar. The nonlinear models describing the dynamics of these two configurations in the lateral/vertical plane are derived. The open-loop characteristics of the two systems are compared by linearizing the nonlinear models about a symmetric hovering equilibrium condition. The closed-loop characteristics of the two systems are compared using nonlinear controllers based on feedback linearization schemes. The performance of the resulting closed-loop systems in carrying out a typical twin-lift mission is evaluated through nonlinear simulation. Also, the effects of helicopter performance degradation and measurement errors on the overall system performance are discussed.[B] Matrix multiplying the control vector in the nonlinear model[B₁] Matrix multiplying the control vector in the linear model[C] Matrix defining vector of variables to be controlled[C₁] Damping matrixC_ijElement of the damping matrix e Parameter used in the linear model = M ₁ h ₁/I ₁=M ₂ h ₂/I ₂,/ft{f} Vector independent of controls in the nonlinear model g Acceleration due to gravity, ft/sec² h₁, h₂Distance of tether attachment point to the center of gravity for helicopters 1 and 2, ft h Parameter used in the linear model, =h ₁=h ₂, ft h Distance between rotor hub and the helicopter center of gravity, ft h h/l H Distance of the load from the spreader bar c.g., ftH₁, H₂Length of tethers 1 and 2, ftI_RMass moment of inertia of spreader bar, slug-ft² I₁, I₂Roll moments of inertia of helicopters 1 and 2, slug-ft² k Non-dimensional hub control moment coefficientK_DDerivative gainsK_IIntegral gainsK_PProportional gains[K_i] Stiffness matrixK_ijElement of the stiffness matrix l Parameter used in the linear model, =H ₁=H ₂, ft L Spreader bar length, ftNomenclature     

Period doubling bifurcation problems in the softening Duffing oscillator with nonstationary excitation

The Duffing oscillators are widely used to mathematically model a variety of engineering and physical systems. A computational analysis has been initiated to explore the effects of nonstationary excitations on the response of the softening Duffing oscillator in the region of the parameter space where the period doubling sequences occur. Significant differences between the stationary and nonstationary responses have been uncovered: (i) the stationary transitions from T to 2T, from 2T to 4T ... etc. branches at the stationary period doubling bifurcations are smooth, in nonstationary cases they exhibit jumps to the near stationary branches at the values of the control parameters greater than those for the stationary; this phenomenon is called penetration (delay or memory). The lengths of the penetrations is being compressed to zero with the increasing number of the iterations. (ii) The stationary and nonstationary responses eventually settle on different limit motions, the nonstationary has modulated components. (iii) The jumps appearing in the stationary bifurcation diagram at 2T from the upper to the lower branches of the (x, f) and (x, ), i.e., (displacement-forcing amplitude) and (displacement-forcing frequency), diagrams have been replaced by continuous transition in the nonstationary diagram climinating thus the discontinuity. Apart from these differences, some specific characteristic nonstationary responses have been observed not encountered in the stationary cases: (iv) the appearance of the window in the nonstationary limit bifurcation diagrams. (v) The nonstationary limit motions located on the upper (lower) branches of the (x, f) or (x, ) diagrams expanded rapidly to the lower (upper) branches. (vi) The stationary and nonstationary bifurcation diagrams are extremely sensitive to the initial conditions, manifested by the mirror reflections, called the flipflop phenomenon. (vii) The nonstationary limit motion has been characterized by a complex phase portrait, the appearance of the Cantor-like set of the limit motion bifurcation plot, and continuous spectral density. For the purpose of comparison, a stationary period doubling sequence T, 2T,..., 2 ⁿ T,... stationary limit motion, _ST which is known to be chaotic has been determined. A far reaching observation has been made in the process of this study: the determination of the nonstationary bifurcations, their branches and limit motions, has been independent of the calculations of the stationary ones, indicating, thus, the existence of an independent class of nonstationary (time varying) dynamics.     

On the Hamiltonian Hopf Bifurcations in the 3D Hénon–Heiles Family

An axially symmetric perturbed isotropic harmonic oscillator undergoes several bifurcations as the parameter adjusting the relative strength of the two terms in the cubic potential is varied. We show that three of these bifurcations are Hamiltonian Hopf bifurcations. To this end we analyse an appropriately chosen normal form. It turns out that the linear behaviour is not that of a typical Hamiltonian Hopf bifurcation as the eigen-values completely vanish at the bifurcation. However, the nonlinear structure is that of a Hamiltonian Hopf bifurcation. The result is obtained by formulating geometric criteria involving the normalized Hamiltonian and the reduced phase space.     

Stability of limit cycles and bifurcations of generalized van der Pol oscillators:

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.     

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.     

On the viscosity-temperature behavior of polymer melts

Based on the free volume concept and the equation by Doolittle, an empirical equation is offered for the flow activation energy, E ^*, for polymer melts for the range of over 150°C above glass transition temperature, T _g. This E ^* represents the temperature coefficient of viscosity for the Newtonian region which is also equal to the value measured at constant shear stress for non-Newtonian flow. Data show that the E ^* of linear polymers approaches a constant value for a temperature range above T _g+150°C. Data on 17 polymers are correlated. The proposed equation for this region predicts the E ^* of polymer melts from the volume expansion coefficient, _l, above T _g and also from the T _g.Correlations have also been developed between E ^* and _l and between E ^* and T _g by simplifying the equation by use of the Simha-Boyer expression. A polymer having a lower _l or higher T _g generally has a higher E ^*. However, more satisfactory results are obtained by calculating E ^* from both _l and T _g. The E ^* calculated is found to agree with measurements within the experimental precision of about ±1 Kcal/mole.The effects of polymer composition, molecular weight, branching and microstructure on E ^* are also discussed. These factors influence E ^* in the way in which they effect _l and T _g.     

Collinear Central Configurations and Singular Surfaces in the Mass Space

For a given m=(m₁,...,m_n)(R⁺)ⁿ, let p and q(R³)ⁿ be two central configurations for m. Then we call p and q equivalent and write pq if they differ by an SO(3) rotation followed by a scalar multiplication as well as by a permutation of bodies. Denote by L(n,m) the set of equivalent classes of n-body collinear central configurations in R³ for any given mass vector m=(m₁,...,m_n)(R⁺)ⁿ. The main discovery in this paper is the existence of a union H₃ of three non-empty algebraic surfaces in the mass half space (m₁,m₂–m₁,m₃–m₂)R⁺×R² besides the planes generated by equal masses, which decreases the number of collinear central configurations. The union H₃ in R⁺×R ² is explicitly constructed by three 6-degree homogeneous polynomials in three variables such that, for any mass vector m=(m₁,m₂,m₃)(R⁺)³, ^# L(3,m)=3, if m₁, m₂, and m₃ are mutually distinct and (m₁,m₂–m₁,m₃–m₂)H₃, ^# L(3,m)=2, if m₁, m₂, and m₃ are mutually distinct and (m₁,m₂–m₁,m₃–m₂)H₃, ^# L(3,m)=2, if two of m₁, m₂, and m₃ are equal but not the third, ^# L(3,m)=1, if m₁=m₂=m₃. We give also a sharp upper bound on ^#L(n,m) for any positive mass vector m(R⁺)ⁿ.     

Numerical analysis of three-dimensional flow past a high-speed electric train with turbulent boundary layer separation

An approach to the numerical study of three-dimensional flow past a high-speed electric train is considered, including the case of turbulent boundary layer separation. The method of viscous-inviscid interaction is used to compute the aerodynamic characteristics. The results of calculating the 3D flow past two configurations of a high-speed electric train, taking into account the close proximity of the ground surface, are presented for a train speedV =300 km/h and the per meter Reynolds number Re=5.6·10⁶ m^–1. One of these configurations is shown to have the advantage of separationless flow past a front locomotive and less intense diffuser separations on the afterbody of a rear locomotive. A local separation zone on the sides of the front locomotive's nose was detected in one of the cases considered.Translated from Izvestiya Rossiiskoi Akademii Nauk, Mekhanika Zhidkosti i Gaza, No. 5, pp. 89–97, September–October, 1993.     

Examination of the flow around a hot-wire probe using particle image velocimetry

Previous work has shown that the k ² term in the effective cooling velocity equation for inclined hot-wires can become negative under certain probe configurations and wire length-to-diameter ratios. It was hypothesised that this was due to a downwash component of velocity along the wire when prong interference effects were expected to be minimal. Direct measurements of the flow around a typical hot-wire probe using digital particle image velocimetry have shown that this downwash velocity component does exist, leading to negative values of k ² as calculated from the angle of deviation from the free stream.List of symbols d diameter of hot-wire mm - k factor in equation for effective dimensionless velocity for inclined hot-wire - l length of hot-wire mm - Q effective velocity mm/s - U free stream velocity mm/s - angle between free stream and degrees wire normal - angle through which flow is degrees deflected at working section of wire     

States of classical statistical mechanical systems of infinitely many particles. I

We study a general mathematical model of a classical system of infinitely many point particles. The space X of infinite particle configurations is equipped with a natural topology as well as a measurable structure related to it. It is also connected with a family {X _A } of local spaces of finite configurations indexed by bounded open sets A in the one-particle space E. A theorem analogous to Kolmogoroff's fundamental theorem for stochastic processes is proved, according to which a consistent family { _A } of local probability measures _A defined on the X _A gives rise to a unique probability measure on X. We also study the problem of integral representation for positive linear forms defined over some linear space of real functions on X. We prove that a positive linear form F(f), defined for functions f in the class C+P, admits a uniquely determined integral representation F(f)= f () d, where is a probability measure over X.     

Dynamic interaction of various beams with the underlying soil – finite and infinite, half-space and Winkler models

Various beams lying on the elastic half-space and subjected to a harmonic load are analyzed by a double numerical integration in wavenumber domain. The compliances of the beam–soil systems are presented for a wide frequency range and for a number of realistic parameter sets. Generally, the soil stiffness G has a strong influence on the low-frequency beam compliance whereas the beam parameters EI and m^′ are more important for the high-frequency compliance. An important parameter is the elastic length l=(EI/G)^1/4 of the beam–soil system. Around the corresponding frequency ω_l=v_S/l, the wave velocity of the combined beam–soil system changes from the Rayleigh wave v_R≈v_S to the bending wave velocity v_B and the combined beam–soil wave has typically a strong damping. The interaction frequency ω_l is found not far from the characteristic frequency ω₀=(G/m^′)^1/2 where an amplification compared to the static compliance is observed for special parameter constellations. In contrast, real foundation beams show no resonance effects as they are highly damped by the radiation into the soil. At medium and high frequencies, asymptotes for the compliance of the beam–soil system are found, u/P(ρv_Paiω)^−3/4 in case of the dominating damping and u/P(−m^′ω²)^−3/4 for high frequencies. The low-frequency compliance of the coupled beam–soil system can be approximated by u/P1/Gl, but it also depends weakly on the width a of the foundation. All numerical results of different beam–soil systems are evaluated to yield a unique relation u/P₀=f(a/l). The integral transform method is also applied to ballasted and slab tracks of railway lines, showing the influence of train speed on the deformation of the track beam. The presented results of infinite beams on half-space are compared with results of finite beams and with infinite beams on a Winkler support. Approximating Winkler parameters are given for realistic foundation-soil systems which are useful when vehicle-track interaction is analyzed for the prediction of railway induced vibration.     

Stability and exponential convergence for the Boltzmann equation

We prove existence, uniqueness and stability for solutions of the nonlinear Boltzmann equation in a periodic box in the case when the initial data are sufficiently close to a spatially homogeneous function. The results are given for a range of spaces, including L ¹, and extend previous results in L for the non-homogeneous equation, as well as the more developed L ^p -theory for the spatially homogeneous Boltzmann equation.We also give new L -estimates for the spatially homogeneous equation in the case of Maxwellian interactions.     

An improved method for the measurement of strong flow birefringence

An improved procedure is proposed to solve the problem of measuring strong birefringence on flowing samples which have a retardation larger than /2 The procedure can be applied to devices which are based on the modulation of the polarization vector with a high-speed rotating half-wave plate. It uses two optical configurations which have different responses to the retardation. By combining the information obtained from the two configurations, one can easily and accurately determine the actual retardation of the sample, irrespective of the quadrant in which the measured retardation is located, provided the sample is isotropic. This technique can also overcome the problem of large errors occurring in the vicinity of = (2m+1)/2 due to the limited sensitivity around these values of . The use of the technique is illustrated with measurements on an isotropic solution of poly(benzyl glutamate) in m-cresol, which is strongly birefringent during shear flow.     

Numerical computation of high Rayleigh number natural convection and prediction of hot radiator induced room air motion

The two-dimensional stationary turbulent buoyant flow and heat transfer in a cavity at high Rayleigh numbers was computed numerically. The k– turbulence model was used. The time-averaged equations for momentum, energy and continuity, which are coupled to the turbulence equations, were solved using a finite difference formulation. In order to validate the computer code, a comparison exercise was carried out. The test results are in good agreement with the internationally accepted benchmark solution. Grid-refinement shows the necessity of a very fine grid at high Rayleigh numbers with especially small grid-distances in the near-wall region. The computed boundary layer velocity profiles are in excellent agreement with available experimental data. The local heat transfer in the turbulent part of the boundary layers is predicted 20% too high. Computations were carried out for the natural convective flow in a room induced by a hot radiator and a cold window. Various radiator configurations and types of thermal boundary conditions were applied including thermal radiation interaction between surfaces.Nomenclature a thermal diffusivity (m²/s) - C constant in _t expression - D cavity dimensions (m) - g acceleration of gravity (m/s²) - G _k production/destruction of k by buoyancy (kg/ms³) - h enthalpy (J/kg) - IX index of grid point - k turbulent kinetic energy (m²/s²) - m dimensionless stratification parameter - Nu overall Nusselt number - Nu _y local Nusselt number - NX total number of grid points - p pressure (N/m²) - P _k production of k by shear stress (kg/ms³) - Q heat flux through wall (W/m) - Ra overall Rayleigh number - Ra _y local Rayleigh number - Re _t turbulent Reynolds number - S source term in -equation (kg/ms⁴) - S source term for - T _c, T _h temperatures of cold and hot walls (K) - T _s (y) stratification temperature on vertical mid-line (K) - T ₀ mean cavity temperature (K) - u, v horizontal and vertical velocity components (m/s) - u ₀ Brunt-Vaisälä velocity scale (m/s) - x, y horizontal and vertical coordinates (m) - non-linearity parameter for grid - coefficient of thermal expansion (l/K) - jet angle (°) - diffusivity for - S dissipation rate for turbulent kinetic energy (m²/s³) - variable to be solved - thermal conductivity (W/mK) - , _t kinematic and eddy viscosities (m²/s) - stream function (kg/ms) - density (kg/m³) - _k, , _t constants in k– model     

Controllable infinitesimal deformations in homogeneously deformed compressible elastic materials

A controllable static deformation is a deformation that may be maintained in all materials of a given class under the action of surface forces alone. For compressible, homogeneous, isotropic elastic materials the only controllable deformations are homogeneous. However, it is known that there are solutions of the static equations of finite elasticity, linearized about a finite homogeneous deformation, which do not correspond to homogeneous deformations. These approximate solutions are investigated here. Three cases arise, depending on whether none, two, or three of the basic principal stretches are equal.Nomenclature A arbitrary vector potential - a ₁, a ₂, a ₃ bounding coordinates of body - B, B _ij left Cauchy-Green tensor - C, C _ijpq elasticity tensor - c, c ₁, c ₂, c ₃ arbitrary constants - N ₀, N ₁, N ₂ elastic response functions - n vector normal to surface of body - T ₁, T ₂, T ₃ surface tractions - t ₁, t ₂, t ₃ surface tractions - t, t _ij Cauchy stress tensor - t ⁰, t _ij ⁰ Cauchy stress corresponding to basic homogeneous deformation - u, u _i infinitesimal displacement from basic homogeneous deformation - X, X _i position vector in reference state - x, x _i position vector - arbitrary function - _ij Kronecker delta - , ₁, ₂, ₃ principal stretches - arbitrary function - arbitrary function - arbitrary function - I, II, III principal invariants of B