Numerical simulations of complex yield-stress fluid flows

Viscoplasticity is characterized by a yield stress, below which the materials will not deform and above which they will deform and flow according to different constitutive relations. Viscoplastic models include the Bingham plastic, the Herschel-Bulkley model and the Casson model. All of these ideal models are discontinuous. Analytical solutions exist for such models in simple flows. For general flow fields, it is necessary to develop numerical techniques to track down yielded/unyielded regions. This can be avoided by introducing into the models a regularization parameter, which facilitates the solution process and produces virtually the same results as the ideal models by the right choice of its value. This work reviews several benchmark problems of viscoplastic flows, such as entry and exit flows from dies, flows around a sphere and a bubble and squeeze flows. Examples are also given for typical processing flows of viscoplastic materials, where the extent and shape of the yielded/unyielded regions are clearly shown. The above-mentioned viscoplastic models leave undetermined the stress and elastic deformation in the solid region. Moreover, deviations have been reported between predictions with these models and experiments for flows around particles using Carbopol, one of the very often used and heretofore widely accepted as a simple “viscoplastic” fluid. These have been partially remedied in very recent studies using the elastoviscoplastic models proposed by Saramito.     

Objective flow classification parameters and their use in general steady flows

Three quantitative flow classification parameters have been studied in the context of Tanner and Huilgol’s suggestion of strong and weak flows. Seen in this context, the different types of streamlines possible for general 3-D flows furnish no indication with respect to the flow strength. This is in total contrast to 2-D flows, where the type of the streamline and the strength of the flow go hand in hand. Astarita’s [J Non-Newton Fluid Mech, 6:69–76, 1979] flow classification parameter takes care of this fact and, if properly generalized, can be applied to more general flows: Two other flow classification parameters also have their basis in homogeneous 2-D flows, but their generalization leads, for general flows, to nonuniqueness and other unacceptable results. For 3-D flows, none of the parameters can quantitatively be used in general, and additional parameters, with their basis outside the 2-D flow regime, seem to be called for.

On the steady solution of linear advection problems in steady recirculating flows

Numerical modelling of non-Newtonian flows typically involves the coupling between equations of motion characterized by an elliptic behaviour, and the fluid constitutive equation, which is an advection equation linked to the fluid history. In this paper we prove that linear steady advection problems in steady recirculating flows have only one solution when the kinematics differs from a rigid motion. We also give a numerical procedure to determine this steady solution. We will describe this numerical procedure for two linear models the first will be the SFRT flow model and the second will be a simplified linear formulation of the Pom–Pom viscoelastic model.     

On the fiber orientation in steady recirculating flows involving short fibers suspensions

In this work we present a new numerical strategy to treat the 3D Fokker–Planck equation in steady recirculating flows. This strategy combines some ideas of the method of particles, with a more original treatment of the periodicity condition, which characterizes the steady solution of the FP equation in steady recirculating flows, as usually encountered in some rheometric devices. Using this numerical technique the fiber orientation distribution can be computed accurately in any steady recirculating flow. The simulation results can be used to identify some rheological parameters of the suspension, using an inverse technique, as well as to analyze the validity of some simplified models widely used, which require a closure relation. Thus, in this paper several closure relations of the fourth-order orientation tensor will be discussed in the context of a numerical example involving a steady recirculating flow.     

The general and basic boundary value problems of plane steady jet flows and the corresponding nonlinear equations

The general boundary value problem, including known plane steady jet flows of an ideal incompressible fluid, is formulated. The simplest problem retaining all the specific features of the general problem, known as the basic problem, is separated from the general problem. The solution of the basic problem is reduced to solving a non-linear integro-differential equation and also to solving nonlinear integral equations. Examples of flows whose determination is reduced to' solving the basic problem are cited.     

Some geometric acoustic problems of one-dimensional unsteady and two-dimensional steady flows

The behavior of discontinuities (weak shocks) of the parameters of a disturbed flow and their interaction with the discontinuities of the basic flow in the geometric acoustics approximation, when the variation of the intensity of such shocks along the characteristics or the bicharacteristics is described by ordinary differential equations, has been investigated by many authors. Thus, Keller [1] considered the case when the undisturbed flow is three-dimensional and steady, and the external inputs do not depend on the flow parameters. An analogous study was made by Bazer and Fleischman for the MGD isentropic flow of an ideal conducting medium [2], while Lugovtsov [3] studied the three-dimensional steady flow of a gas of finite conductivity for small magnetic Reynolds numbers and no electric field. Several studies (for example, [4]) have considered the behavior of discontinuities of the solutions from the general positions of the theory of hyperbolic systems of quasilinear equations. Finally, the interaction of weak shocks (or the equivalent continuous disturbances) with shock waves was studied in [5–11].In what follows we consider one-dimensional (with plane, cylindrical, and spherical waves) and quasi-one-dimensional unsteady flows, and also plane and axisymmetric steady flows. Two problems are investigated: the variation of the intensity of weak shocks in the presence of inputs which depend on the stream parameters, and the interaction of weak shocks with strong discontinuities which differ from contact (tangential) discontinuities.The thermodynamic properties of the gas are considered arbitrary. We note that the resulting formulas for the interaction coefficients of the weak and strong discontinuities are also valid for nonequilibrium flow.     

Collision of two plane free streams with division of one of them into two flows

The collision of two streams of ideal incompressible fluid with the division of one stream into two flows in opposite directions along the boundary of the other is considered. The second stream is shifted from its initial direction due to the collision. The ratio of the total pressure heads of the separated and shifted streams is 1. The flow regions corresponding to each of these streams map conformally onto the first and fourth quadrants of the complex plane so that the streamline separating the streams corresponds to the real semiaxis in both cases. By using the dynamic and kinematic conditions for this streamline, nonlinear relations are found linking the boundary values of the logarithms of the complex velocity in each of the two flow regions. These analytic functions are computed approximately by linearization at small values of .Translated from Izvestiya Akademii Nauk SSSR, Mekhanika Zhidkosti i Gaza, No. 1, pp. 3–7, January–February, 1975.     

On intrinsic properties of steady gas flows

Summary Considering the geometric theory of triply orthogonal spatial curves, the basic equations governing a steady gas flow are transformed into the intrinsic form and the results obtained are:(1) The pressure is uniform along the binormal to the stream line and the radius of curvature varies as the square of the velocity along it, for the baratropic fluids.(2) Acceleration is irrotational field when the fluid is compressible but baratropic or incompressible, in which case the relations existing between the flow quantities, curvature and torsions of the curves under consideration are obtained.(3) Considering incompressible flows, it is observed that either velocity in magnitude is uniform or the vorticity lies in the normal plane, in which case the stream lines are orthogonal to the vortex lines.Stream lines are observed to be either right circular helices or circles or straight lines.If the stream lines are not straight then the torsions of the binormal congruences and stream lines are equal.(4) The compatibility conditions of Berker¹) are transformed into intrinsic form, involving the curvatures and torsions of the above curves.     

Non-elliptic elastic materials and the modeling of dissipative mechanical behavior: an example

In the context of a special problem, this paper investigates the possibility of modeling dissipative mechanical response in solids on the basis of the equilibrium theory of finite elasticity for materials that may lose ellipticity at large strains. Quasi-static motions for such materials are in general dissipative if the associated equilibrium fields involve discontinuous displacement gradients. For the problem treated, consideration of such deformations is shown to lead naturally to an internal variable formalism similar to those used to describe macroscopic plastic behavior arising from microstructural effects. For quasi-static motions which are maximally dissipative in a specified sense, this formalism leads to a mechanical response which resembles that associated with the pseudo-elastic effect in shape-memory alloys.     

A generalized anisotropic hardening rule based on the Mroz multi-yield-surface model for pressure insensitive and sensitive materials

In this paper, a generalized anisotropic hardening rule based on the Mroz multi-yield-surface model for pressure insensitive and sensitive materials is derived. The evolution equation for the active yield surface with reference to the memory yield surface is obtained by considering the continuous expansion of the active yield surface during the unloading/reloading process. The incremental constitutive relation based on the associated flow rule is then derived for a general yield function for pressure insensitive and sensitive materials. Detailed incremental constitutive relations for materials based on the Mises yield function, the Hill quadratic anisotropic yield function and the Drucker–Prager yield function are derived as the special cases. The closed-form solutions for one-dimensional stress–plastic strain curves are also derived and plotted for materials under cyclic loading conditions based on the three yield functions. In addition, the closed-form solutions for one-dimensional stress–plastic strain curves for materials based on the isotropic Cazacu–Barlat yield function under cyclic loading conditions are summarized and presented. For materials based on the Mises and the Hill anisotropic yield functions, the stress–plastic strain curves show closed hysteresis loops under uniaxial cyclic loading conditions and the Masing hypothesis is applicable. For materials based on the Drucker–Prager and Cazacu–Barlat yield functions, the stress–plastic strain curves do not close and show the ratcheting effect under uniaxial cyclic loading conditions. The ratcheting effect is due to different strain ranges for a given stress range for the unloading and reloading processes. With these closed-form solutions, the important effects of the yield surface geometry on the cyclic plastic behavior due to the pressure-sensitive yielding or the unsymmetric behavior in tension and compression can be shown unambiguously. The closed form solutions for the Drucker–Prager and Cazacu–Barlat yield functions with the associated flow rule also suggest that a more general anisotropic hardening theory needs to be developed to address the ratcheting effects for a given stress range.     

Drag difference between steady and shocked gas flows passing through a porous body

Experimental and numerical investigations of gas flows through porous materials have been carried out. We have investigated steady and unsteady processes occurring when the gas flow interacts with porous materials. Densities and porosities of the four open-cell-type polyurethane foams which were investigated are kg/m and , with the foams having different structures. Experiments were conducted to determine the steady drag coefficient of the porous material at low Reynolds numbers, evaluated from the pressure drop. The Forchheimer equation was applied to determine the drag. Values of permeability coefficients () in the Forchheimer equation were estimated by comparing computed and experimental results. Results show that the drag coefficient is largely affected by the internal structure of the foam, and has almost no effect on the stress history, while the value of dominates the stress history variation. Differences of 1000 times exist between the steady flow and unsteady shock tube flow values. Received 15 May 1998/ Accepted 15 March 1999     

Conservative calculations of non-isentropic transonic flows

The classical potential formulation of inviscid transonic flows is modified to account for non-isentropic effects. The density is determined in terms of the speed as well as the pressure, which in turn is calculated from a second-order mixed-type equation derived via differentiating the momentum equations. The present model differs in general from the exact inviscid Euler equations since the flow is assumed irrotational. On the other hand, since the shocks are not isentropic, they are weaker and are placed further upstream compared to the classical potential solution. Furthermore, the streamline leaving the aerofoil does not necessarily bisect the trailing edge. Results for the present conservative calculations are presented for non-lifting and lifting aerofoils at subsonic and transonic speeds and compared to potential and Euler solutions.     

On calculation of hydrodynamic forces for steady flows in unbounded domains

This note addresses the question why the “impulse formula”, often employed to compute hydrodynamic forces in vortex-dominated time-dependent flows, is not applicable to steady flows in unbounded domains. By analyzing the asymptotic structure of steady and unsteady flow solutions in unbounded domains, it is demonstrated that one assumption made in the derivation of the impulse formula is in fact not satisfied in the steady case. This result also highlights the special character of steady flows in unbounded domains.     

Vorticity in plane parallel, axially symmetrical or spherical flows of viscous fluid

For viscous (barotropic or incompressible) fluids it is shown that, if the vorticity and the viscous force are orthogonal, vortex lines are convected by a vector field which fits with the velocity field when viscosity vanishes (extension of Helmholtz theorem); it is also found that energy remains constant along the field lines of this vector field (extension of Bernoulli theorem).If, moreover, vorticity and velocity are orthogonal too, the magnitude of the vorticity then behaves as the density of a fluid which flows along streamsheets according to this very same vector field. These properties are mainly encountered for plane parallel flows, axially symmetrical flows, spherical flows, but also for some other miscellaneous flow geometries such as unidirectional or radial flows. The set of the former three flows can even be characterized by these properties; that enhances this set of important flow geometries, avails a general view on vorticity behavior, and explains the great simplicity of vorticity equations in these cases. Numerous examples and comments are given for illustrating.     

The Giesekus fluid in ω–D form for steady two-dimensional flows

Using the fact that for simple fluids the most general constitutive equation in constant stretch history flows for the extra stress tensor τ is known in an explicit form, the Giesekus fluid model is cast into this (ω–D) form for two-dimensional flows. The three material functions needed to characterize τ are listed. The explicit results for simple shear and planar elongation reveal that the parameter α should be restricted to values less than 0.5. It is demonstrated that in this explicit form the constitutive equation is free from thermodynamic objections and can thus be used as a starting point for numerical calculations of general, but steady, two-dimensional flows. Received: 9 November 1998 Accepted: 20 May 1999     

Nonisothermal instability of high-velocity elastoplastic flows

An important feature of the high-velocity deformation of solids is the localization of deformation, one of the causes of which may be the nonisothermal instability of plastic flow [1–6]. In connection with the intensive development of high-velocity technology in the treatment of materials, the investigation of the criteria for nonisothermal stability of processes of plastic deformation is of fundamental interest, since in certain cases they determine the optimum technological regimes [5]. The critical values of deformation velocities, above which the effects of thermal instability becomes decisive in the process of deformation of solids, are estimated by semiempirical methods in [1]. The non-boundary-value problem of the criteria for nonisothermal instability is analyzed in [2] for the point of view of flow stability in the so-called coupled formulation. The latter means that the heat-conduction equation is added to the basic equations determining the dynamics of an elastoplastic medium. The problem is solved in [6] in an analogous formulation, but for flow averaged over the spatial coordinate. The solution of the boundary-value problem for one-dimensional flow in this formulation is given in the present paper.Translated from Zhurnal Prikladnoi Mekhaniki i Tekhnicheskoi Fiziki, No. 3, pp. 133–138, May–June, 1986.     

Rate sensitivity of mixed mode interface toughness of dissimilar metallic materials: Studied at steady state

Crack propagation in metallic materials produces plastic dissipation when material in front for the crack tip enters the active plastic zone traveling with the tip, and later ends up being part of the residual plastic strain wake. Thus, the macroscopic work required to advance the crack is typically much larger than the work needed in the near tip fracture process. For rate sensitive materials, the amount of plastic dissipation typically depends on the rate at which the material is deformed. A dependency on the crack velocity should therefore be expected. The objective of this paper is to study the macroscopic toughness of crack advance along an interface joining two dissimilar rate dependent materials, characterized by an elastic-viscoplastic material model that approaches the response of a J₂-flow material in the rate independent limit. The emphasis here is on the rate sensitivity of the macroscopic fracture toughness under mixed Mode I/II loading. Moreover, special cases of joined similar rate dependent materials, as well as dissimilar materials where one substrate remains either elastic or approaches the rate independent limit is also included. The numerical analysis is carried out using the SSV model [Suo, Z., Shih, C., Varias, A., 1993. A theory for cleavage cracking in the presence of plastic flow. Acta Metall. Mater. 41, 1551–1557] embedded in a steady state finite element formulation, here assuming plane strain conditions and small-scale yielding. Results are presented for a wide range of material parameters, including noteworthy observations of a characteristic crack velocity at which the macroscopic toughness becomes independent of the material rate sensitivity. The potential of this phenomenon is elaborated on from a modeling point of view.     

Simulation of polymeric flows in the injection moulding process

Recent progress in the simulation of polymeric flows of two key problems in the injection moulding process, carried out by a team at Cornell University, is briefly described. For the filling of cooled thin cavities, the fluid is characterized by a power-law viscosity with exponential temperature dependence, and interaction between the transient thermal boundary-layer and the core flow in a domain with moving boundary is essential. The earlier procedure of Hieber and Shen is modified in two aspects: a boundary-integral formulation replaces the finite-element treatment of the pressure, and an ‘energy integral’ approach is used for the transient temperature. The second problem is the steady visco-elastic flow in the juncture region where sudden changes of the geometry and large strain rates occur. The constitutive equation is postulated according to the Leonov model. The main features in the numerical implementation are: integration along a streamline to determine the elastic deformation tensors for a given velocity field, and finite-element treatment (in time-dependent form) of the pressure and fields for given stresses. In an example where the contraction ratio is 7:1, results for nominal Deborah number exceeding 100 show no numerical instability. (However, for this problem, the true Weissenberg number, i.e. the ratio of local first-normal-stress difference to shear stress turns out to be generally O(10).) The predictions also correlate very well with experimental birefringence measurements.     

Gauge principle and variational formulation for flows of an ideal fluid

A gauge principle is applied to mass flows of an ideal compressible fluid subject to Galilei transformation. A free-field Lagrangian defined at the outset is invariant with respeet to global SO(3) gauge transformations as well as Galilei transformations. The action principle leads to the equation of potential flows under constraint of a continuity equation. However, the irrotational flow is not invariant with respect to local SO(3) gauge transformations. According to the gauge principle, a gauge-covariant derivative is defined by introducing a new gauge field. Galilei invariance of the derivative requires the gauge field to coincide with the vorticity, i.e. the curl of the velocity field. A full gauge-covariant variational formulation is proposed on the basis of the Hamilton‘‘s principle and an assoicated Lagrangian. By means of an isentropic material variation taking into account individual particle motion, the Euler‘‘s equation of motion is derived for isentropic flows by using the covariant derivative. Noether‘‘s law associated with global SO(3) gauge invariance leads to the conservation of total angular momentum. In addition, the Lagrangian has a local symmetry of particle permutation which results in local conservation law equivalent to the vorticity equation.     

Numerical simulations of cessation flows of a Bingham plastic with the augmented Lagrangian method

The augmented Lagrangian/Uzawa method has been used to study benchmark one-dimensional cessation flow problems of a Bingham fluid, such as the plane Couette flow, and the plane, round, and annular Poiseuille flows. The calculated stopping times agree well with available theoretical upper bounds for the whole range of Bingham numbers and with previous numerical results. The applied method allows for easy determination of the yielded and unyielded regions. The evolution of the rigid zones in these unsteady flows is presented. It is demonstrated that the appearance of an unyielded zone near the wall occurs for any non-zero Bingham number not only in the case of a round tube but also in the case of an annular tube of small radii ratio. The advantages of using the present method instead of regularizing the constitutive equation are also discussed.