Resonant non-linear vibrations in continuous systems—I. Undamped case

A method is presented for obtaining periodic solutions to forced oscillations of non-linear systems governed by equations of the form u_ss?u_yy?εf(u,u,_y,u_yy…,s) = 0. The method is presented by application to the equation u_ss?u_yy?εu²_yu_yy= 0 which governs the vibrations of a soil layer that is free on the top surface and is forced harmonically at the bedrock. It is shown that unlike the ODE case (Duffing equation), the PDE requires an infinite number of periodicity conditions to correctly characterize the resonant region and these conditions lead to an infinite number of branches in the dispersion spectrum. Calculations indicate that these branches tend to an envelope curve. The uniform approach presented by Millmann and Keller is discussed in order to determine in what sense it can be viewed as an effective approximation for the fundamental mode.     

Statistical Analysis and a-priori Modelling of Flame Surface Density Transport in Turbulent Stratified Flames: A Direct Numerical Simulation Study

Statistically planar turbulent premixed and stratified flames for different initial intensities of decaying turbulence have been simulated for global equivalence ratios <???> = 0.7 and <???> = 1.0 using three-dimensional simplified chemistry based Direct Numerical Simulations (DNS). The simulation parameters are chosen such that the thin reaction zones regime combustion is realised in all cases and a random bi-modal distribution of equivalence ratio ? is introduced in the unburned gas ahead of the flame to account for the mixture inhomogeneity for stratified flames. The modelling of the unclosed terms (i.e. the turbulent transport term T ₁, the tangential strain rate term T ₂, the propagation term T ₃, and the curvature term T ₄) of the generalised FSD transport equation has been addressed in the context of RANS simulations. It has been found that the turbulent transport term T ₁ remains small in comparison to the leading order contributions of the tangential strain rate and curvature terms (i.e. T ₂ and T ₄, respectively) in the globally stoichiometric cases, but T ₁ begins to play a more important role in the globally fuel-lean cases. The strain rate term T ₂ remains positive throughout the flame brush and acts as a leading order source term for all the flames considered in this analysis. It is has been found that the magnitude of T ₂ decreases with decreasing root-mean-square velocity fluctuations u ^′ (<???>) for a given value of <???> (u ^′). The contribution of the propagation term T ₃ remains generally positive towards the unburned gas side of the flame brush but assumes generally negative values towards the burned gas side of the flame brush. Moreover, whilst the order of magnitude of the propagation term T ₃ is comparable in all cases, T ₃ remains small in comparison to the leading order contributors (i.e. T ₂ and T ₄) in the globally stoichiometric cases however it plays a more important role in the globally fuel-lean cases. The curvature term T ₄ acts as a leading order sink term in all cases except towards the unburned gas side of the flame brush in low u ^′ globally stoichiometric (i.e. <???> = 1.0) flames. Furthermore, it has been demonstrated that the magnitude of T ₄ decreases with decreasing u ^′ (<???>) for a given value of <???> (u ^′). Appropriate model expressions have been identified for T ₁, T ₂, T ₃ and T ₄ based on an a-priori analysis of the DNS data.     

Les solutions periodiques et les points fixes de l'application translation dans une equation differentielle ordinaire

We consider an ordinary differential equation E, x(t) = F(t, x(t)), with time periodic right hand side, with period T. The translation mapping Θ is the one which transforms an initial point Y at time t₀ into the value at time t₀ + T of the solution of E with initial conditions (t₀, Y). It is known that the solution with initial conditions (t₀, Z) is periodic with period T if and only if Z is a fixed point of Θ. In this paper the Newton's method is applied to locate the fixed points of the translation mapping Θ.     

Canonical ensemble for static elastic structures with random microstructures

A constant-(N,V,T,θ) ensemble is proposed to describe the elastostatics of random solid structures. Within the harmonic approximation, the energy of such a solid structure is the sum of a thermal and a strain component without mutual interaction. Systems in this ensemble thus draw energies from two separate baths: one thermal and the other mechanical. A mechanical entropy and an effective temperature (θ) can then be defined on the same rigorous basis as the thermal entropy and the Kelvin temperature (T). This ensemble approach can be used to calculate the properties of solid structures sharing similar microstructural randomness.     

Calibration of a temperature dissipation probe in decaying grid turbulence

A four cold-wire probe, which allows all three components of the temperature dissipation rate ? _θ to be measured, is “calibrated” in decaying grid turbulence, where <? _θ >, the mean value of ? _θ , can be determined accurately from the decay of the temperature variance <θ ²>. The probe yields values of the three components of <? _θ > which are in reasonable agreement with local isotropy, in the range x ₁/M?50. The pdfs and spectra of the three temperature derivatives also satisfy local isotropy reasonably well.     

Transient heat and mass transfer in a natural circulation loop

Comprehensive work has been performed by theoretical and numerical methods in order to study the steady state, transient and stability characteristics of a double diffusive natural circulation loop. It was found that the behavior of the flow in the system depends on the initial conditions and on the location of the state in the seven-parameter space of the thermal and saline Rayleigh numbers,Ra _T ,Ra _S , the modified Prandtl and Schmidt numbers,Pr, Sc, the dimensionless heat and mass transfer coefficients,H _T ,H _S , and the “aspect ratio” (between the height and width) of the loop, γ. Numerical results are presented here, showing the flow in each of the five regions formed in the stability chart. The steady state solutions include convection (constant velocity flow), conduction (no-flow) and periodic with constant amplitude and frequency. Two main new results were obtained: long term periodic oscillations where the amplitude is not symmetric around the conduction solution, and an overshoot of the velocity in transients before reaching the stable convection solutions. In the monotonic instability region of the conduction solution, convection solutions (constant velocity flow) develop, and in the global stability region the flow decays to the conduction solution (no flow), regardless of the initial conditions.     

Separation zones in the flow near a small-angle convex corner

On the basis of an asymptotic analysis of the Navier-Stokes system of equations for large Reynolds numbers (Re → ∞), the plane incompressible fluid flow near a surface having a convex corner with a small angle 2θ* is investigated. It is shown that for θ* = O(Re^?1/4), in addition to the known solution that describes a separated flow completely localized in a thin “viscous” sublayer of the interaction region near the corner point, another solution corresponding to a flow with a developed separation zone is possible. For θ ₀ = Re^1/4 θ* = O(1), the longitudinal dimension of this zone varies from finite values up to values of the order of Re^?3/8. The nonuniqueness of the solution is established on a certain range of variation of the parameter θ ₀. The dependence of the drag coefficient on the angle θ* is found.     

Exact solutions and Painlevé analysis of a new (2+1)-dimensional generalized KdV equation

The new (2+1)-dimensional generalized KdV equation which exists the bilinear form is mainly discussed. We prove that the equation does not admit the Painlevé property even by taking the arbitrary constant a=0. However, this result is different from Radha and Lakshmanan??s work. In addition, based on Hirota bilinear method, periodic wave solutions in terms of Riemann theta function and rational solutions are derived, respectively. The asymptotic properties of the periodic wave solutions are analyzed in detail.     

Front Quenching in the G-equation Model Induced by Straining of Cellular Flow

We study homogenization of the G-equation with a flow straining term (or the strain G-equation) in two dimensional periodic cellular flow. The strain G-equation is a highly non-coercive and non-convex level set Hamilton–Jacobi equation. The main objective is to investigate how the flow induced straining (the nonconvex term) influences front propagation as the flow intensity A increases. Three distinct regimes are identified. When A is below the critical level, homogenization holds and the turbulent flame speed s _T (effective Hamiltonian) is well-defined for any periodic flow with small divergence and is enhanced by the cellular flow as s _T ≧ O(A/log A). In the second regime where A is slightly above the critical value, homogenization breaks down, and s _T is not well-defined along any direction. Solutions become a mixture of a fast moving part and a stagnant part. When A is sufficiently large, the whole flame front ceases to propagate forward due to the flow induced straining. In particular, along directions p = (±1, 0) and (0, ±1), s _T is well-defined again with a value of zero (trapping). A partial homogenization result is also proved. If we consider a similar but relatively simpler Hamiltonian, the trapping occurs along all directions. The analysis is based on the two-player differential game representation of solutions, selection of game strategies and trapping regions, and construction of connecting trajectories.     

Thermal activation relations

The temperature-dependence of the viscosity η of a liquid can be described by means of an empirical equation suggested as early as 1921 byVogel: η=A exp.B/(T?T ₀), which, besides the temperatureT, contains three adjustable structuredependent parametersA, B andT ₀. An equation of the same form is suited for describing the temperature dependence of the most probable relaxation timeτ _m in polymers, and can be easily transformed into the socalled WLF-equation, because the empirical parametersB andT ₀ are simple functions of the WLF parameters. In many cases, the temperature-dependence of such quantities as mechanical or dielectric relaxation times, viscosities, diffusion coefficients etc, is regarded as a consequence of a thermal activation process ruled by a temperature-dependentGibbs Free Energy of activationΔG. On the basis of this concept both parametersB andT ₀, can be related to the more fundamental high temperature limiting value of the activation energyΔH and the parameterT ₀ to the temperature at whichΔG becomes infinite. The parameterA is normally taken to be temperature-independent; theoretically, however, it may just as well be temperature-dependent. Although it is difficult to decide from experimental evidence whether this is the case (accurate values over a large temperature range must be available), one should realize that analysis done by means of a temperature-independent pre-exponential factor gives a set of values for the parameters which differ greatly from those found when e.g.A is set proportional toT ^?1 orT ^?2. Literature values for the viscosities ofn-paraffins have been statistically analysed. It is concluded that η in theVogel-equation had better be replaced byη T ^1.2/?, where? is the density of the liquid. Values are given for the modifiedVogel, WLF and activation parameters of the paraffins and ethylenecopolymers, those for the latter being obtained from our own dielectricτ _m measurements. By analogy,τ _m would then have to be replaced byτ _m T ^2.2. However with our dielectric data, such a refinement would not give any great improvement.     

An Alternative Approach to Study Bifurcation from a Limit Cycle in Periodically Perturbed Autonomous Systems

The goal of this paper is to present a new method to prove bifurcation of a branch of asymptotically stable periodic solutions of a T-periodically perturbed autonomous system from a T-periodic limit cycle of the autonomous unperturbed system. The method is based on a linear scaling of the state variables to convert, under suitable conditions, the singular Poincaré map (with two singularity conditions) associated to the perturbed autonomous system into an equivalent non-singular equation to which the classical implicit function theorem applies directly. As a result we obtain the existence of a unique branch of T-periodic solutions (usually found for bifurcations of co-dimension 2) as well as a relevant property of the spectrum of their derivatives. Finally, by a suitable representation formula of the classical Malkin bifurcation function, we show that our conditions are equivalent to the existence of a non-degenerate simple zero of the Malkin function. The novelty of the method is that it permits to solve the problem without explicit reduction of the dimension of the state space as it is usually done in the literature by the Lyapunov–Schmidt method.     

Statistical Analysis of Cross Scalar Dissipation Rate Transport in Turbulent Partially Premixed Flames: A Direct Numerical Simulation Study

Statistically planar turbulent partially premixed flames for different initial intensities of decaying turbulence have been simulated for global equivalence ratios <????> = 0.7 and <????> = 1.0 using three-dimensional simplified chemistry based Direct Numerical Simulations (DNS). The simulation parameters are chosen such that the combustion situation belongs to the thin reaction zones regime and a random bi-modal distribution of equivalence ratio ?? is introduced in the unburned gas ahead of the flame to account for mixture inhomogeneity. The DNS data has been used to analyse the statistical behaviour of the transport of the cross-scalar dissipation rate based on the fuel mass fraction Y _F and the mixture fraction ?? fluctuations $\,\tilde{\varepsilon}_{Y\xi}={\overline{\rho D\nabla Y_{F}^{\prime\prime}.\nabla \xi^{\prime\prime}} } \big/ {\bar {\rho }}$ (where $\bar{q}$ , $\tilde{q}={\overline{\rho q} } \big/ {\bar {\rho }}$ and $q^{\prime\prime} =q-\tilde {q}$ are Reynolds average, Favre mean and Favre fluctuation of a general quantity q) in the context of Reynolds Averaged Navier?CStokes simulations where ?? is the gas density and D is the gas diffusivity. The statistical behaviours of the unclosed terms in the $\tilde{\varepsilon }_{Y\xi } $ transport equation originating from turbulent transport T ₁, density variation T ₂, scalar?Cturbulence interaction T ₃, chemical reaction rate T ₄ and the molecular dissipation rate D ₂ have been analysed in detail. It has been observed that the contributions of T ₂, T ₃, T ₄ and D ₂ play important roles in the $\tilde{\varepsilon }_{Y\xi } $ transport for the globally stoichiometric cases, but in the globally fuel-lean cases the contributions of T ₂ and T ₄ become relatively weaker in comparison to the contributions of T ₃ and D ₂. The term T ₁ remains small in comparison to the leading order contributions of T ₃ and D ₂ for all cases, but the contribution of T ₁ plays a more important role in the low Damköhler combustion cases. The term T ₂ behaves as a sink term towards the unburned gas side but becomes a source term towards the burned gas side. The scalar?Cturbulence interaction term T ₃ has been found to be generally positive throughout the flame brush, but in globally stoichiometric cases the contribution of T ₃ becomes negative in regions of intense heat release. The combined contribution of (T ₄ ?C D ₂) remains mostly as a sink in all cases studied here. Models are proposed for the unclosed terms of the $\tilde{\varepsilon }_{Y\xi } $ transport equation in the context of Reynolds Averaged Navier?CStokes simulations, which are shown to satisfactorily predict the corresponding quantities extracted from the DNS data for all cases.     

Periodic solutions induced by an upright position of small oscillations of a sleeping symmetrical gyrostat

The aim of this paper is to provide sufficient conditions for the existence of periodic solutions emerging from an upright position of small oscillations of a sleeping symmetrical gyrostat with equations of motion being α and β parameters satisfying Δ=α ²?4β>0 and $\beta-\frac{\alpha^{2}}{2}\pm \frac{\alpha \sqrt{\varDelta }}{2}<0$ , ε a small parameter and, F ₁ and F ₂ smooth periodic maps in the variable t in resonance p:q with some of the periodic solutions of the system for ε=0, where p and q are positive integers relatively prime. The main tool used is the averaging theory.     

Existence of Time-Periodic Solutions for a Magnetoelastic System in Bounded Domains

We investigate the existence of periodic solutions for a semilinear (nonlinearly coupled) magnetoelastic system in bounded, simply connected, three-dimensional domains with boundaries of class C ². The mathematical model includes a nonlinear mechanical dissipation like ρ(u′)=|u′| ^p u′ and a periodic forcing function of period T. We prove the existence of T-periodic weak solutions when p∈[3,4] (p=0 being a simpler case). In the corresponding two-dimensional case, the existence result holds under the assumption that p≥2.     

Determination of source parameter in parabolic equations

The authors consider the problem of finding u=u(x, t) and p=p(t) which satisfy u = Lu + p(t) + F(x, t, u, _x, p(t)) in Q _T=Ω×(0, T], u(x, 0)=ø(x), x∈Ω, u(x, t)=g(x, t) on ?Ω×(0, T] and either ∫_G(t) Φ(x,t)u(x,t)dx = E(t), 0 ? t ? T or u(x₀, t)=E(t), 0≤t≤T, where Ω?R ⁿ is a bounded domain with smooth boundary ?Ω, x ₀∈Ω, L is a linear elliptic operator, G(t)?Ω, and F, ø, g, and E are known functions. For each of the two problems stated above, we demonstrate the existence, unicity and continuous dependence upon the data. Some considerations on the numerical solution for these two inverse problems are presented with examples.     

Bounded solutions of some nonlinear elliptic equations in cylindrical domains

The existence of a (unique) solution of the second-order semilinear elliptic equation $$\sum\limits_{i,j = 0}^n {a_{ij} (x)u_{x_i x_j } + f(\nabla u,u,x) = 0}$$ withx=(x ₀,x ₁,?,x _n )?(s ₀, ∞)× Ω′, for a bounded domainΩ′, together with the additional conditions $$\begin{array}{*{20}c} {u(x) = 0for(x_1 ,x_2 ,...,x_n ) \in \partial \Omega '} \\ {u(x) = \varphi (x_1 ,x_2 ,...,x_n )forx_0 = s_0 } \\ {|u(x)|globallybounded} \\ \end{array}$$ is shown to be a well-posed problem under some sign and growth restrictions off and its partial derivatives. It can be seen as an initial value problem, with initial value?, in the spaceC ₀ ⁰ $(\overline {\Omega '} )$ and satisfying the strong order-preserving property. In the case thata _ij andf do not depend onx ₀ or are periodic inx ₀, it is shown that the corresponding dynamical system has a compact global attractor. Also, conditions onf are given under which all the solutions tend to zero asx ₀ tends to infinity. Proofs are strongly based on maximum and comparison techniques.     

A generalization of the method of averaging for systems with two time scales

In this paper we shall consider systems of the form x = ? f(t, ?t, x, y, ?), y = g(t,?t, x, y,?), where x and y are vectors of finite dimensions, f and g are assumed to be bounded for all t, and ? is a real parameter. Sufficient conditions are obtained for the existence of certain solutions which are bounded for all t. These solutions are shown to approach special solutions of a derived simpler averaged system of equations as ? → 0. Moreover, it is shown that there exists only one such bounded solution in the neighborhood of each special solution. In the special case when y is not present, it is shown that if a special solution is stable, solutions starting in nonlocal neighborhoods of this special solution approach the bounded solutions adjacent to it as t → ∞. These results generalize most of the existing work for systems of the type discussed here. Finally, we employ our results to study some problems of physical importance.