Analysis of two-dimensional cavity flow by finite elements

The variational principle in terms of stream function ψ for free surface gravity flow is discussed by the formulation of first-order variation in a variable domain. Because of different transversal conditions adopted, there are four forms of variational principle in terms of ψ.A n air-gilled cavity flow with given discharge and total energy is then analysed by finite element method. At the end of the cavity, the free stream line is tangent to a short fictitious plate of given length, which joins the fixed boundary at on angle to be determined. The condition that the free stream line should be tangent to the fixed boundary at the point of separation makes the solution unique.Finally curves giving the cavity length as a function of the Fraude number, cavity pressure and channel bottom slope are presented.     

Analysis of two-dimensional,finite amplitude wave propagation by time marching methods

Two-dimensional, finite-amplitude wave propagation in an inviscid, subsonic, perfect gas medium is analysed by explicit finite-difference methods. A two-step, Lax-Wendroff method and the single-step, Lax-Friedrichs method are used. A prescribed propagating velocity or pressure disturbance is applied along a single row of grid points normal to the stream direction and results in a 'forced' outflow boundary. The inflow boundary is placed far from outflow by utilizing a streamwise expanding grid and uniform inflow is imposed. Side boundaries are spatially periodic. The numerical solutions are compared with analytical small-perturbation solutions; higher-order effects arising from non-linearities are revealed by Fourier analysis. Solutions which closely approached a periodic state were obtained. The Lax-Wendroff method combined with the expanding grid is shown to be accurate and stable, the Lax-Friedrichs scheme produced highly damped solutions.     

Variational irreversible thermodynamics of physical-chemical solids with finite deformation

A new thermodynamics of open thermochemical systems and a variational principle of virtual dissipation are applied to the finite deformation of a solid coupled to thermomolecular diffusion and chemical reactions. A variational derivation is obtained of the field differential equations as well as Lagrangian equations with generalized coordinates. New formulas for the affinity and a new definition of the chemical potential are presented. An outline is given of an unusually large field of applications, such as active transport in biological systems, finite element methods, plastic properties as analogous to chemical reactions, phase changes and recrystalization, porous solids, heredity and initially stressed solids. A new and unified insight is thus provided in highly diversified problems.     

Multiscale modeling of the dynamics of solids at finite temperature

We develop a general multiscale method for coupling atomistic and continuum simulations using the framework of the heterogeneous multiscale method (HMM). Both the atomistic and the continuum models are formulated in the form of conservation laws of mass, momentum and energy. A macroscale solver, here the finite volume scheme, is used everywhere on a macrogrid; whenever necessary the macroscale fluxes are computed using the microscale model, which is in turn constrained by the local macrostate of the system, e.g. the deformation gradient tensor, the mean velocity and the local temperature. We discuss how these constraints can be imposed in the form of boundary conditions. When isolated defects are present, we develop an additional strategy for defect tracking. This method naturally decouples the atomistic time scales from the continuum time scale. Applications to shock propagation, thermal expansion, phase boundary and twin boundary dynamics are presented.     

A point assembly method for stress analysis for two-dimensional solids

A novel point assembly method (PAM) is presented for stress analysis for two-dimensional solids. In the present method, the boundaries of the problem domain are represented by a set of discrete points, and the domain itself is represented by properly scattered points. The displacement in the influence triangular areas of a point is interpolated by the displacements at the point and pairs of surrounding points using shape functions. The shape functions used in this work are obtained in the same way as those of a triangular element in the conventional finite element method (FEM). A variational (weak) form of the equilibrium equation is used to produce a set of system equations. These equations are assembled for all the points in the domain, and solved for the displacement field. Stresses and strains at a point are then computed using the displacements obtained for the point and pairs of the surrounding points. A PAM program with an automatic point-searching algorithm has been developed in fortran. Patch tests and convergence studies have been carried out to verify the convergence of the present method and program. Examples are also presented to demonstrate the efficiency and accuracy of the present method compared with analytical solutions as well as the conventional FEM solutions.     

Description of finite deformation of solids in a reference configuration

Variational principles of equilibrium processes of deformation and heat conduction are formulated. The variational relations are written for the initial configuration of solids and can be used without restrictions on the strain. A system of equations for a coupled boundary-value problem for isotropic and anisotropic solids is presented, and initial and boundary conditions are formulated. Results of solving problems of finite deformation of initially cylindrical solids are given.     

Coupled fields of two-dimensional anisotropic magneto-electro-elastic solids with an elliptical inclusion

ＩｎｔｒｏｄｕｃｔｉｏｎＤｅｔｅｒｍｉｎａｔｉｏｎｏｆｐｅｒｔｕｒｂｅｄｆｉｅｌｄｓｉｎｄｕｃｅｄｂｙｉｎｃｌｕｓｉｏｎｓｈａｓｂｅｅｎａｎｉｍｐｏｒｔａｎｔｔｏｐｉｃｆｏｒｓｔｕｄｙｉｎｇｔｈｅｐｈｙｓｉｃａｌｂｅｈａｖｉｏｒｓｏｆａｄｖａｎｃｅｄｃｏｍｐｏｓｉｔｅｍａｔｅｒｉａｌｓ.Ｆｒｏｍｔｈｅｐｏｉｎｔｏｆｖｉｅｗｏｆａｐｐｌｉｃａｔｉｏｎｓ,ｔｈｅａｄｖａｎｃｅｄｃｏｍｐｏｓｉｔｅｍａｔｅｒｉａｌｓｃａｎｂｅｄｉｖｉｄｅｄｉｎｔｏｔｗｏｃａｔｅｇｏｒｉｅｓ:ｏｎｅｉｓｔｈｅｓｔｒｕｃｔｕ…     

Some two-dimensional flows at finite magnetic reynolds numbers

Flows at finite magnetic Reynolds numbers are characterized by a strong effect of the induced magnetic fields on the stream. In the present paper we determine the current distribution and estimate the influence of the Lorentz force component perpendicular to the stream in a two-dimensional channel with electrodes. We also estimate the influence of nonuniformities of the velocity in the stream path of an incompressible fluid when the characteristic magnetic Reynolds numbers     

Towards the solution of finite plane-strain problems for compressible elastic solids

A new approach to the solution of finite plane-strain problems for compressible Isotropie elastic solids is considered. The general problem is formulated in terms of a pair of deformation invariants different from those normally used, enabling the components of (nominal) stress to be expressed in terms of four functions, two of which are rotations associated with the deformation. Moreover, the inverse constitutive law can be written in a simple form involving the same two rotations, and this allows the problem to be formulated in a dual fashion.For particular choices of strain-energy function of the elastic material solutions are found in which the governing differential equations partially decouple, and the theory is then illustrated by simple examples. It is also shown how this part of the analysis is related to the work of F. John on harmonic materials.Detailed consideration is given to the problem of a circular cylindrical annulus whose inner surface is fixed and whose outer surface is subjected to a circular shear stress. We note, in particular, that material circles concentric with the annulus and near its surface decrease in radius whatever the form of constitutive law within the given class. Whether the volume of the material constituting the annulus increases or decreases depends on the form of law and the magnitude of the applied shear stress.     

Path-independentJ-integral and its dual form in elastic-plastic solids with finite displacements

In this paper, based on energy variational principles of elastic-plastic solids, the path-independentJ-integral and its dual form in elastic-plastic solids with finite displacements are presented. Whose testification is given there after.     

Mixed finite element for two-dimensional incompressible convective Brinkman-Forchheimer equations

In this work, the two-dimensional convective Brinkman-Forchheimer equations are considered. The well-posedness for the variational problem and its mixed finite element approximation is established, and the error estimates based on the conforming approximation are obtained. For the computation, a one-step Newton(or semi-Newton)iteration algorithm initialized using a fixed-point iteration is proposed. Finally, numerical experiments using a Taylor-Hood mixed element built on a structured or unstructured triangular mesh are implemented. The numerical results obtained using the algorithm are compared with the analytic data, and are shown to be in very good agreement. Moreover,the lid-driven problem at Reynolds numbers of 100 and 400 is considered and analyzed.     

A consistent finite element model for the two-dimensional continuum

Summary The finite element approximation to the continuum problem is examined from the viewpoint of the principle of virtual work. It is shown that the usual nodal equilibrium equations for triangular elements are a consistent consequence of a piecewise constant strain field, thus guaranteeing that many results of general continuum theory can be directly applied to the finite element model, and also clarifying the relation between the two models.

---

Übersicht Das Verfahren, ein Kontinuum durch finite Elemente anzunähern wird vom Standpunkt des Prinzips der virtuellen Arbeiten untersucht. Es wird gezeigt, daß die üblichen Knotenpunktsgleichungen für dreieckförmige Elemente eine Folge des stückweise konstanten Verformungsfeldes sind. Auf diese Weise wird sichergestellt, daß viele Ergebnisse der allgemeinen Kontinuumstheorie unmittelbar auf das aus endlichen Elementen aufgebaute Modell übertragen werden können. Gleichzeitig werden die Beziehungen zwischen beiden Modellen geklärt.

---

---

Dedicated to Professor Dr. H. Ziegler on the occasion of his 60th birthday.

---

This research was sponsored by the National Science Foundation, Grant GK 10549.     

PARTITION OF UNITY FINITE ELEMENT METHOD FOR SHORT WAVE PROPAGATION IN SOLIDS

IntroductionIt is known that standard finite element procedure is unable to simulate the wavepropagation with high oscillations or gradients in space in the media with reasonableefficiency and accuracy due to the nature of polynomial interpolation approxi…     

Adaptive unstructured finite element method for two-dimensional detonation simulations

A non-equilibrium reacting flow methodology has been added to a conservative, monotonic, compressible flow solver to allow numerical simulations of gas detonations. This flow solver incorporates unstructured dynamically adaptive meshes with the Finite Element Method – Flux Corrected Transport (FEM-FCT) scheme, which has shown excellent predictive capability of various non-reacting compressible flows. A two-step induction parameter model was used to model the combustion of the gas phase coupled with an energy release equation which was simulated with a point implicit finite element scheme. This combustion model was then applied to a two-dimensional detonation test case of a hypothetical fuel:oxygen mixture. The detonation simulation yielded two transverse waves which continued to propagate. This feature and the detonation shock speed mean and fluctuations were found to be grid-independent based on a resolution of about twenty elements within the average induction length. The resolution was efficiently achieved with the unstructured dynamically adaptive finite elements, which were three orders of magnitude less in number then required for uniform discretization. Received 26 August 1996 / Accepted 31 March 1997     

BIEM analysis of dynamically loaded anti-plane cracks in graded piezoelectric finite solids

Anti-plane cracks in finite functionally graded piezoelectric solids under time-harmonic loading are studied via a non-hypersingular traction based boundary integral equation method (BIEM). The formulation allows for a quadratic variation of the material properties in two directions. The boundary integral equation (BIE) system is treated by using the frequency dependent fundamental solution based on Radon transforms. Its numerical solution provides the displacements and tractions on the external boundary as well as the crack opening displacements from which the mechanical stress intensity factor (SIF) and the electrical displacement intensity factor (EDIF) are determined. Several examples for single and multiple straight and curved cracks demonstrate the applicability of the method and show the influence of the different system parameters.     

A finite element solution of unsteady two-dimensional flow in cascades

A theory is presented for unsteady two-dimensional potential transonic flow in cascades of compressor and turbine blades using a mesh of triangular finite elements. The theory leads to a computer program, FINSUP, which is fast and has moderate storage requirements, so that it can be run on a personal computer. Comparisons with other theories in special cases show that the program is accurate in subsonic flow, and that in supersonic flow, although the wave effects are smeared by the numerical process, the results for overall blade force and moment have acceptable accuracy. The program is useful for engineering assessment of unstalled flutter of actual compressor and turbine blades.     

Stability analysis of finite difference schemes for two-dimensional advection-diffusion problems

This paper develops a stability analysis of second-order, two- and three-time-level difference schemes for the 2D linear diffusion-convection model problem. The corresponding 1D schemes have been extensively analysed in two previous papers by the same author. Most of these 2D schemes obviously generalize 1D schemes, i.e. their stencil only uses the nearest points and defines ‘product difference schemes’; however, the stability results are not always the exact generalization of the 1D stability properties. Moreover, the 1D nonviscous MFTCS scheme may only be generalized if one uses a nine-point scheme. Numerical experiments for different values of the cell Reynolds number allow a comparison to be made between the theoretical and numerical stability limits.