We will start the talk with a presentation of the large eddy simulation approach based on relaxation filtering (LES-RF). This method represents a return-to-first-principles in turbulence simulation, in that the artificial-viscosity paradigm, used by most turbulence models of the past half-century, is replaced by an explicit relaxation filtering. Aeroacoustics is today a central element in many engineering areas as aeronautics, ground transportations or energy production.We will summarized recent work regarding the direct computation of aerodynamic noise, and the use of these simulations to improve our understanding of noise generation by turbulent flows.
The aim of this paper is to discuss the relation between dissipative and fatigue phenomena, using the measure of the dissipated energy. This framework presents a unified analysis to both high and low cycle fatigue based on shakedown theories and dissipated energy and has provided successful lifetime predictions of industrial structuires in both HCF and LCF. The presentation will start with a series of examples lifetime predictions of industrial structures and will then discuss the fatigue phenomena at different length scales (microscopic, mesoscopic and macroscopic) using mainly the concept of shakedown.
Based on research results of ultralightweight porous metallic materials published recently in international journals and in Science in China Series E, this article summarizes our combined interdisciplinary efforts to promote the multifunctional implementation of this new class of materials, including their design, production, properties and applications.
Large eddy simulation method is developed for study of decaying and forced compressible magnetohydrodynamic turbulence. The obtained results of numerical computations for large eddy simulation are compared with the results of direct numerical simulation of three-dimensional compressible magnetohydrodynamic turbulence under various similarity parameters, namely, magnetic Reynolds numbers,
hydrodynamic Reynolds numbers and Mach numbers for polytropic case. The comparison of five subgrid-scale closures of large eddy simulation for magnetohydrodynamic case is made: the Smagorinsky model, the Kolmogorov model, the cross-helicity model, the scale-similarity model and mixed model. The comparison between large eddy simulation and direct numerical simulation is carried out regarding the time evolution of kinetic and magnetic energy, cross helicity, subgrid-scale and molecular dissipations for kinetic and magnetic energy, turbulent intensities and quantities that describe anisotropy of flow, that is, skewness and kurtosis of velocity and magnetic field. It is shown that some proposed subgrid-scale models provide sufficient dissipation of kinetic and magnetic energy, reduce computational efforts and produce adequate results of magnetohydrodynamic turbulent modeling for various values of similarity parameters of flows. We also consider a heat conducting compressible fluid with the use of an energy equation. Application of large eddy simulation approach to heat conducting compressible magnetohydrodynamics is considered. It is shown that novel subgrid-scale terms arise in Favre-filtered equations due to the presence of a magnetic field in the total energy equation. Parameterizations of these extra terms are developed. Computations at various Mach numbers are made for decaying compressible magnetohydrodynamic turbulence. The obtained numerical large eddy simulation results are analyzed on the basis of comparison with results of numerical experiments performed by direct numerical simulation. We present the large-eddy simulation method for studying forced compressible magnetohydrodynamic turbulence as well. The proposed method is based on a linear representation of the driving forces in the momentum conservation equation and the magnetic induction equation. These forces supply the production of kinetic and magnetic energy. The emphasis is placed upon the important, and not investigated, question about the ability of the large-eddy simulation approach to reproduce Kolmogorov and Iroshnikov-Kraichnan scale-invariant spectra in compressible magnetohydrodynamic flows.
Statistical theories of turbulence developed so far dealing with the mean velocity products of various orders are outlined historically and reformulated as the non-equilibrium statistical mechanics of fluid turbulence based on the Lundgren-Monin equations for the multi-point velocity distributions and the cross-independence closure hypothesis proposed by Tatsumi. According to this formalism, the infinite set of the Lundgren-Monin equations is closed as the finite set of equations for the n-point
velocity distributions f(n) (n>=1) , the minimum deterministic set of which being composed of the oneand two-point velocity distributions f and f(2), the latter being represented by the velocity-sum distribution g+ and the velocity-difference distribution g-. As an outstanding result, it is shown that the energy-dissipation rate of turbulence e is expressed in terms of the distribution g- which is mostly contributed from small-scale turbulence, making clear analogy with the "fluctuation-dissipation theorem" of non-equilibrium statistical mechanics. Another remarkable result is the exactness of the present closure in the sense that the closed terms of the Lundgren-Monin equations are exactly equivalent with the unclosed terms of the same equations. This property of the closure is confirmed by the fact that the moment equations derived from the closed equations for the velocity distribution are precisely equivalent with the corresponding equations for the mean velocity products derived from the Navier-Stokes equation directly.
Based on research results of theoretical modeling of electro-acoustic devices by functional materials published in international journals, this article summarizes our efforts to the understanding of novel surface acoustic waves and their applications in electro-acoustic devices. Challenges to researchers from mechanics, electronic and material fields are also pointed out.
Based on the experimental investigations conducted earlier on aluminum alloy foam, the bending stiffness was found to be two times larger than the compression stiffness. Our analysis on the experimental results suggested this to be due to the difference in local deformation of cell when it is under compression and bending loading. The work presented in this paper clarifies this finding by means of computational analysis. Two-dimensional finite element model is developed to explain the reason of the discrepancy. The results successfully exhibit the discrepancy in stiffness; the difference in local deformation of cell was found to be the reason of the discrepancy by a unit-cell analysis on the model. Moreover, the results also reveal the strong dependency of stiffness ratio to the density. In addition, a theoretical approach is also proposed to explain the discrepancy in stiffness.
A numerical method to identify thin films Young's modulus from multiple indentation unloading curves was suggested.
Usual inverse analyses make use of the whole indentation load-displacement curve, and thus require elasto-plastic finite element simulations. Subsequently, all the plastic properties should be identified the same time, leading to heavy computations. However, it was reported that the initial state of the indentation unloading can be considered elastic. For this reason, the proposed method by applying the inverse analysis only to this part of the unloading curve can then rely only on elastic finite element computations.
To ensure the applicability of the procedure, several factors that might affect its accuracy are addressed.
Finally, the method is confronted with experimental indentation data of Polyimide film on Glass/Epoxy substrate.
An unconventional finite-strain hyperelasto-plastic constitutive model based on the subloading surface concept is presented. On the basic kinematic assumption of the multiplicative decomposition of the deformation gradient into elastic and plastic parts, the hyperelastic law and plastic evolution equations are formulated on the intermediate configuration which is chosen to be isoclinic. The nonlinear isotropic and kinematic hardening as well as the non-dissipative part of the plastic velocity gradient are incorporated into the model. The computational treatment of the model is also discussed. An effective numerical algorithm for stress update is developed with the use of tensor exponential-based return mapping. As a specific example, von Mises plasticity model with a neo-Hookean type hyperelastic law and linear/nonlinear kinematic hardening is considered. The basic behavior and capability of the proposed model to large strains are demonstrated through several numerical examples.
The singular corner of the original Cam-clay model represents the virgin consolidation for the model but evaluation of plastic flow at this vertex singularity causes numerical trouble when the associated flow rule is applied. In this study a backward-Euler stress update algorithm is developed for the original Cam-clay model. A constraint function is defined as a plane passing the vertex and the Koiter's associated flow rule was employed to both yield and constraint functions. This method can conveniently handle the vertex without discretizing the yield surface into a finite number of yield loci. The new algorithm was applied for four diversified forms of the yield function to consider the effect of yield function's form on the performance of the algorithm. Numerical results of the proposed algorithm were illustrated in the single step and varying number of sub-steps. Error-maps were successfully generated to evaluate a performance of the algorithm.