Jtam-A4.dvi JOURNAL OF THEORETICAL AND APPLIED MECHANICS 52, 2, pp. 405-415, Warsaw 2014 ON SOME GENERAL SOLUTIONS OF TRANSIENT STOKES AND BRINKMAN EQUATIONS D. Palaniappan Department of Mathematics and Statistics, Texas A&M University, Corpus Christi, Texas, USA e-mail: devanayagam.palaniappan@tamucc.edu General solution representations for velocity and pressure fields describing transient flows at small Reynolds numbers (Stokes flows) and flows obeying Brinkmanmodels are presented. The geometrydependent vector representations emerge from the incompressibility condition and are expressed in terms of just two scalar functions similar to the Papkovich-Neuber and Boussinesq-Galerkin solution type.We provide new formulae connecting our differential re- presentations and other solutions describing unsteady Stokes flow including Lamb’s (1932) general infinite series solution. The unified approach presented here further demonstrates an important link between oscillatory flows and flow through porousmedia usingBrinkman models. It is shown that the solutions of boundary value problems in the latter can be obta- ined in a straightforward fashion, from the results of the former. This simple but surprising analogy is further explained using the properties (mathematical aswell as physical) that are shared by the two different models. The construction of certain physical quantities is also illustrated for spherical and spheroidal inclusions. It is believed that the general solutions presented here will be useful in the computation of multi-particle interactions in transient and Brinkman flows and also in linear elasticity. Keywords: transient Stokes flow, differential representation, Brinkman equations 1. Introduction The linearized viscous flow at low-Reynolds numbers is described by a pair of partial differential equations connecting thevelocitywith thepressurefield.The so called transientStokes equations were solvedmanyyears ago for oscillatorymotions of a sphere along adiameter byStokes (1851). Subsequently, Lamb (1945) treated the problems of periodicmotion in three-dimensions having special relations to spherical surfaces in a more general manner. In particular, Lamb (1945) presented a general solution of the oscillatory Stokes equations – in terms of three linearly independent scalar functions – suitable for spherical boundaries in the form of an infinite series by extending his idea for the steady case. In principle, all the problems of periodic motions involving a spherical boundary can be solved using Lamb’s solution in terms of solid spherical harmonics. On the other hand, spherical geometry provides the most widely used framework for representing small particles and obstacles embedded within a viscous, incompressible fluid characterizing transient creeping flow. Other notable works on this topic include those due to Basset (1888), Mazur and Bedeaux (1974), Felderhof (1978), Lawrence and Weinbaum (1986), Yang and Leal (1991) amongst many others. A brief historical review by Pozrikidis (1994) contains further references on this subject. There are many efficient methods in use to solve problems in low-Reynolds number flow theory such as numerical computation, stream function technique, analytical function methods and a differential representation technique. An important feature of the differential representa- tion approach is that it leads to closed form analytic solutions in a fairly simple manner. The famous representations due toBoussinesq (1885), Papkovich-Neuber (Neuber, 1934) andNaghdi and Hsu (1961) have been used extensively to solve various steady and nonsteady problems in 406 D. Palaniappan elasticity and fluid dynamics. In the interest of producing differential representations similar to Papkovich-Neuber and Boussinesq-Galerkin, a general solution in terms of two scalar functions A and B was proposed earlier in the case of steady-state creeping flow (Palaniappan et al., 1992). For steady flows at small Reynolds numbers, the representation due to Palaniappan et al. (1992) yields a complete general solution (Padmavathi et al., 1998). Such a representation emerges as a result of the incompressibility condition and therefore can serve as a complete ge- neral solution for unsteady creeping flowproblems aswell. Indeed the common incompressibility feature for steady and unsteady flows suggests that the representation given in Palaniappan et al. (1992), Palaniappan (2009) can yield a complete set of basis functions for transient flow pro- blems as well. In the first part of this paper, we discuss the connection between the differential representation given in Palaniappan et al. (1992) and other solutions in the context of transient creeping flow. Specifically, we provide new formulae connecting the differential representation and other solutions describing unsteady viscous flow. In particular, we show that the Lamb’s general solution follows from the differential representation by a suitable choice of the scalar functions. The connections to other representations are briefly discussed. Another differential representation suitable for bounded flows (Palaniappan, 2000, 2009) constrained by plane wall is also given. This general representation is shown to generate solution forms that are suitable for studying oscillatory motions of disks at low Reynolds numbers. In the second part, we demonstrate the link between the oscillatory creeping flow and flow through porous media. The equations postulated by Brinkman (1947) in modeling of porous media have been found widely applicable for high porosity systems. The merits of these equ- ations over Darcy’s equations (modeling systems with low porosity) may be found in the work of Oomes et al. (1970), Neale et al. (1973), Masliah et al. (1987), and Higdon and Kojima (1981). Moreover, the validity of Brinkman’s equations has been justified theoretically by Tam (1969), Saffman (1971), Lundgren (1972), Howells (1974). Furthermore, the problems involving multiparticle interactions in a porous medium (Kim and Russel, 1985) and porous spherical shells (Qin andKaloni, 1993; Bhatt and Sacheti, 1994; Padmavathi and Amaranath, 1996) also use Brinkman’s equations effectively. Some perspectives on convection problems indicating the merits and limitations of equations of porous media may be found in Nield and Bejan (1992). Although the Brinkman model is physically quite different from transient flow model, it has some interesting mathematical similarities as shown here. 2. Solutions of transient creeping flow The linearized Navier Stokes equations describing the motion of a viscous, incompressible fluid are ρ ∂u ∂t = µ∇2u−∇p ∇·u=0 (2.1) where u, p are the velocity and pressure fields for the fluid and µ, ρ are the fluid viscosity and fluid density, respectively. For oscillatory motions, u and p vary as eαt (α is assumed to be imaginary) that is, u=u′eαt, p = p′eαt and therefore equation (2.1)1 may be written as µ(∇2+h2)u′ =∇p′ (2.2) where h2 =−α/ν, ν is the kinematic viscosity. Using (2.1)2 in (2.2) we obtain ∇ 2p′ =0 (2.3) and now (2.2) reduces to (∇2+h2)∇2u′ =0 (2.4) On some general solutions of transient Stokes and Brinkman equations 407 Thus, the pressure in transient flow is harmonic as in the steady case but the Laplacian of the velocity vector u′ satisfies the Helmholtz equation. It may be worthwhile to point out that in the steady-state creeping flow, the Laplacian of velocity vector is harmonic and, hence, it is biharmonic there. It is evident from (2.2) (by operating curl on both sides) that the vorticity vector satisfies the Helmholtz equation. Incorporating these features, Lamb (1945) presented a general solution of (2.2) and (2.3) as u ′ = ∞ ∑ n=−∞ [ 1 h2µ ∇pn+(n+1)ψn−1(hr)∇φn −nψn+1(hr)h 2r2n+1∇ φn r2n+1 +ψn(hr)r×∇χn p′ = pn (2.5) The solid harmonic functions φn, χn arise from the solution of the homogeneous equation and pn is the particular solution of (2.2). The function ψn is related to the Bessel function of a fractional order that is finite at the origin. In the situation where the quantities of interest are finite at infinity, the function ψn is to be replaced by Ψn. Their relations to theBessel functions are as follows ζnψn(ζ)= √ π 2ζ Jn+1 2 (ζ) and ζnΨn(ζ)= (−) n √ π 2ζ J −n−1 2 (ζ) It is worth mentioning here that equation (2.2) may be regarded as the Laplace transform of (2.1)1 with the transformvariable s = h 2.Thenby taking the inverseLaplace transformof (2.5)1, one could obtain a general solution of transient Stokes equations for arbitrary time-dependent flow. 2.1. A general representation for velocity and pressure fields In the theory of isotropic elasticity andhydrodynamics, it is common to assume the solutions of the governingpartial differential equations in termsof auxiliary scalar functions, often referred to as differential representations. For instance, in linear elasticity, the displacement vector may be represented in terms of scalar functions as in Boussinesq (1885), Neuber (1934), Naghdi and Hsu(1961) knownasBoussinesq,Papkovich-Neuber andNaghdi-Hsudifferential representations, respectively. This approach is also followed and applied to many problems in hydrodynamics including inviscid (Milne-Thomson, 1968) and viscous (Stokes, 1851; Palaniappan et al., 1992; Dassios andVafeas, 2004) flows. The crucial point in this technique is that the governing vector differential equations in each physicalmodel reduce to solving scalar differential equations for the auxiliary functions. Inmany circumstances, the boundary conditions also turn out to be simpler to apply with the scalar functions technique. The differential representation approach is also applicable for time-dependent viscous flow models. Since the velocity vector (u and hence u′) in transient flow is divergence-free, a suitable representation for this quantity is u ′ = curl curl(rA)+ curl(rB) (2.6) where r= xi+yj+zk and A, B are the scalar functions.The representation (2.6)was originally proposed for steady-state incompressible flows (Palaniappan et al., 1992) and has been shown to yield a complete general solution for creeping flows (Padmavathi et al., 1998). This type of differential representation has also been employed inFeng et al. (1998) to investigate the general motion of a circular disk in a Brinkman medium (a problem that is mathematically equivalent to oscillatory Stokes flow discussed in Section 3 and 4 below). 408 D. Palaniappan Substitution of (2.6) into (2.2) yields the pressure p′ = µ ∂ ∂r [r(∇2+h2)A] (2.7) if ∇ 2(∇2+h2)A =0 (∇2+h2)B =0 (2.8) Thus, vector equation (2.1)1 for u reduces to solving scalar equations (2.8) for the functions A and B. The scalar function A can be decomposed into A = A1+A2 where ∇ 2A1 = 0 and (∇2+h2)A2 =0. It is interesting to note that the function A1 belongs to the kernel of ∇ 2 (the Laplace operator) and the functions A2 and B belong to the kernel of ∇ 2+h2 (the Helmholtz operator). We now provide connection formulae by which we can transform any solution of oscillatory flow from Lamb’s solution to differential representation form (2.6) and vice-versa. Following the steady case approach, we take A = ∞ ∑ −∞ [ 1 h2µ pn n+1 +(2n+1)ψn(hr)φn ] B = ∞ ∑ −∞ ψn(hr)χn (2.9) Substitution of (2.9) into (2.6) and (2.7) yields Lamb’s solution (2.5) after the use of some sim- ple vector identities. Therefore, expressions (2.9) provide the formulae connecting differential representation (2.6) and Lamb’s general solution for transient flows. Connections to other so- lution representations, including the cartesian tensor solution due to Felderhof (1978), may be established in a similar manner. For flows with axial symmetry, the solution can be found using a single scalar function commonly known as the Stokes stream function. In (2.6), if we take (after expanding the curl operator) ∂A ∂θ = ψ(r,θ) r sinθ B =0 (2.10) ((r,θ) refer to spherical polar coordinates) we then obtain the axisymmetric representation of the solution for oscillatory flow. In this case, the Stokes stream function ψ(r,θ) satisfies the fourth-order scalar equation D2(D2+h2)ψ =0 D2 = ∂2 ∂r2 + 1−η2 r2 ∂2 ∂η2 η =cosθ The stream function ψ(r,θ) belongs to the kernel of D2(D2+h2) and can be decomposed into two functions of which one belongs to the kernel of D2 and the other to the kernel of (D2+h2). Note that u′ and p′ are treated as time-independent quantities and so are the functions A and B. But it should be remembered that they are in fact related to the original velocity and presseure fields u and p through u = u′eαt and p = p′eαt. For this reason, we retain the name oscillatory flow in the following examples, although, the time-dependence is not explicitly mentioned. 2.1.1. Examples (i) Uniform oscillatory flow. For a uniformoscillatory flow along the x-axis, the scalar functions A and B in (2.6) are A(r,θ,φ) = U 2 r sinθcosφ B =0 U > 0 (2.11) On some general solutions of transient Stokes and Brinkman equations 409 The corresponding functions in Lamb’s general solution (2.5) are obtained using connection formulae (2.9) as p1 = Ur sinθcosφ pn =0 for n ­ 2 φn = χn =0 for all n (2.12) For the oscillatory flow symmetrical about z-axis (axis of symmetry), (2.10) yields the stream function ψ(r,θ) = U 2 r2 sin2θ, (2.13) which was used by Stokes (1851) to study the oscillation of a sphere along its diameter. (ii) Oscillatory rotlet along z-axis. For the flow resulting from a rotlet located at the origin and whose axis along z-direction, one has A =0 B = cosθ r2 e−R(R+1) (2.14) where R = hr. Comparison of (2.14) with (2.9) yields χ1 = cosθ r2 ψ1 =e −R(R+1) pn = φn =0 (2.15) which are the functions inLamb’s solution representing the rotlet flow.The function ψ1 in (2.15) can also be expressed in terms of modified Bessel’s functions. In general, the transient flowfields are completely determined through the scalar functions A and B given in differential representation (2.6). This general representation yields a set of basis functions suitable for problems involving spherical boundaries. Itmay be possible to analyze the utility of (2.6) for other body shapes including spheroids and ellipsoids. 2.2. A general solution representation for a plane wall For flows in a semi-infinite domain, which is constricted by a plane wall, amore appropriate complete general differential representation is (Palaniappan, 2000) u= curl curl(êzA)+ curl(êzB) p = p0+µ ∂ ∂z [(∇2+h2)A] (2.16) where ∇2(∇2 +h2)A = 0, (∇2 +h2)B = 0, and êz is the unit vector along z-direction. The image solutions for oscillating point singularities located in front of a plane wall (Pozrikidis, 1991) can be constructed using representation (2.16). Differential representation (2.16) is also suitable for solving boundary value problems con- cerning the oscillatory motions of circular disks in a viscous fluid. To see this, we first write the velocity components and pressure in the cylindrical coordinates (ρ,θ,z) as uρ = ∂2A ∂ρ∂z + 1 ρ ∂B ∂θ uθ = 1 ρ ∂2A ∂θ∂z − ∂B ∂ρ uz =− (∂2A ∂ρ2 + 1 ρ ∂A ∂ρ + 1 ρ2 ∂2A ∂θ2 ) (2.17) with p = p0+µ ∂ ∂z (∇2+h2)A 410 D. Palaniappan By taking A(ρ,θ,z) = ψ(ρ,z) ρ cosθ B(ρ,θ,z)= χ(ρ,z) ρ sinθ we obtain uρ = [ ∂ ∂ρ (1 ρ ∂ψ ∂z ) + χ ρ2 ] cosθ uθ =− [ 1 ρ2 ∂ψ ∂z + ∂ ∂ρ (χ ρ )] sinθ uz =− ∂ ∂ρ (1 ρ ∂ψ ∂ρ ) cosθ (2.18) and p = p0+µ cosθ ρ ∂ ∂z (D2+h2)ψ These are precisely the functional forms assumed by Zhang and Stone (1998) to solve edgewise translation of a disk. Likewise, the other particular functional forms that are zero and first order harmonics in θ used in Zhang and Stone (1998) can be deduced from (2.16). Clearly, the flow problems involving higher harmonics in θ do not belong to the class of solutions discussed in Zhang and Stone (1998) and require a more generic approach. Needless to say that such problems can be treated by the use of the general differential representation given in (2.16). Another general solution incylindrical coordinates is givenbyHappelandBrenner (1983) in their monograph.Their solution iswell suited for steadyStokes flowproblems involving infinitely long circular cylinders. One can also find explicit relations connecting the differential representation given in (2.16) and Happel and Brenner’s general solution. It should be pointed out that differential representations (2.6) and (2.16)1 emerge from the incompressibility condition and differ by the vectors r and ez. Note that for spherical bounda- ries r represents a normal vector to the sphere and for planar surfaces ez is a vector normal to the xy-plane. It is evident that the general solution representation is geometry dependent. Indeed, the boundary conditions in terms of the scalar functions A and B become considerably simpler. Thismakes our solution representationsmore effective andmay be preferred over other solution forms for solving boundary value problems in transient Stokes flows. Below, we discuss the solutions given in (2.6) and (2.16)1 in the context of Brinkman’s equations whichmodel the flow through porousmedia. 3. Solutions of Brinkman model equations Now the equations modeling the flow through a porousmedium proposed by Brinkman are µ∇2v− µ1 k v=∇p ∇·v=0 (3.1) where v, p and µ have the usual meaning, k is the permeability of the porous medium and µ1 is the effective viscosity. Equation (3.1)1 may conveniently be written as µ(∇2−λ2)v=∇p (3.2) where λ2 = µ1/(kµ). Note that equations (2.2) and (3.2) have the same structure (except for the sign of λ2 and h2) and hence reveal the fact that the two different models are mathemati- cally equivalent. It is straightforward to see (by doing the same operations as before) that the pressure is harmonic, and the Laplacian of velocity and vorticity satisfy Helmholtz’s equation in the Brinkman medium as in the case of transient flow. Thus, the equations of porous media On some general solutions of transient Stokes and Brinkman equations 411 and oscillatory flows share some of the physical properties as well. The sign difference in (2.2) and (3.2) merely indicates the choice of the Bessel functions that are having complex and real arguments respectively (Watson, 1952). Hence, the functions Jn are to be replaced by In in the porous media model in order to obtain the general solution of Brinkman’s equations. We therefore make the following observations: • the velocity field given in (2.6) in terms of the scalar functions A and B is a suitable differential representation for Brinkman’s equations. • Lamb’s general solution given in (2.5) provides an alternative representation of the solu- tion for Brinkman’s equations in terms of the scalar functions φn, pn and χn (all time- independent in this case) with h2 =−λ2. The above observations add to the list ofmathematical similarities between the transient Stokes equations and Brinkman equations which have been noted earlier in the literature (Kim and Russel, 1985; Feng et al., 1998; Pozrikidis, 1991). Similarities between the twomodels have also been observed for themotion of non-spherical objects (Feng et al., 1998; ZhangandStone, 1998). Therefore, it appears thatwith an appropriate interpretation, it is possible to obtain solution for one model from the other. Some specific cases illustrating the above observations are discussed in the next Section. Itmaybenoted that for planarboundaries in theBrinkmanmedium, the solution representa- tion given in (2.16) provides a complete general solution. Indeed, this differential representation can be used to findwall images due to point singularities located in the vicinity of porous slabs. The different solution forms assumed in Feng et al. (1998) for studying motion of disks in the Brinkman medium can be deduced from (2.16) in the same way as explained in the previous Section. It follows that the scalar representation of solutions of the transient Stokes equations and Brinkman equations are exactly the same as in the case of the steady Stokes flow which follows from the incompressibility condition. It is therefore unnecessary to discuss such repre- sentations for these twomodels separately.Moreover, forms similar to (2.6) and (2.16)1 can also be utilized in classical linear elasticity and elastodynamics as well. 4. Further illustration We now turn our attention to justify the foregoing observations by considering some special situations.We restrict ourselves to a class of boundary value problemswhere the velocity vector vanishes at the surface of abody that is immersed in aflow (no-slip or stick boundarycondition). Although the solution of theBrinkman equations can be obtained directly fromLamb’s solution (Eqs. (2.5)), little care should be taken in extracting the physical quantities. Since unsteady flows exhibit acceleration, an additional term is always present for instance in the force or drag. It could be easily isolated from the force and the remaining part excluding the time factor gives the expression for the force acting on a particle submerged in aBrinkmanmedium.We illustrate these facts in the examples considered below.For convenience, we denote the quantities for time- dependent flows and porousmedia (Brinkman’smodel) with suffixed t and B, respectively and λ as a common parameter. But it should be understood that λ assumes its respective values in two different models. 4.1. Force on a sphere Consider the Stokes problem of an oscillating spherical pendulum in a viscous fluid. The velocity and pressure fields for this problem are given elsewhere (Stokes, 1851; Lamb, 1945). The force acting on the blob of radius a oscillating about x-axis is 412 D. Palaniappan Ft = [ −6πµUa ( 1+λ+ λ2 3 ) + 4 3 πµUaλ2 ] eαt (4.1) We note that the second term on the right hand side of (4.1) is purely due to inertia which is absent in the Brinkmanmodel. Dropping the second term together with the time-dependent factor eαt in (4.1), one obtains FB =−6πµUa ( 1+λ+ λ2 3 ) (4.2) The above expression yields the drag on a test sphere in a Brinkman medium and agrees with the result given inTam (1969). As said before, the suffixes denote the time-dependent flows (the explicit time-dependence is now clearly seen in (4.1)) and the Brinkman medium, respectively. We have used λ as a common parameter in both the cases and it takes different values in the respective phases. 4.2. Force on a spheroid Now we consider a slightly oblate spheroid immersed in an oscillating flow. The expression for the force acting on the spheroid is (Lawrence andWeinbaum, 1986) Ft = { −6πµUa [( 1+λ+ λ2 3 ) + 4 5 ǫ(1+2λ+λ2) + 2 175 ǫ2 ( 1+58λ+53λ2+ 4λ2 3+3λ+λ2 )] + 4 3 πµUaλ2(1+ ǫ+ ǫ2) } eαt (4.3) where ǫ is the departure from the spherical shape. For an oblate spheroid ǫ > 0 and for a prolate spheroid ǫ < 0. Dropping the last term and together with eαt as in the previous case, we obtain the new result FB =−6πµUa [( 1+λ+ λ2 3 ) + 4 5 ǫ(1+2λ+λ2)+ 2 175 ǫ2 ( 1+58λ+53λ2+ 4λ2 3+3λ+λ2 )] (4.4) The above expression yields the hydrodynamic force acting on a spheroid submerged in aBrink- man medium. When the permeability is large i.e., when λ is small, we recover the Stokes resistance for a slightly oblate spheroid suspended in a steady flow (Happel andBrenner, 1983). For small permeability, equation (4.4) is dominatedby λ2 term. In this case, the O(λ) correction to (4.4) is important because it describes the growth of the boundary layer at the body surface. It may also be noted that when ǫ =0, (4.4) reduces to that of a perfect sphere. Finally, the signdifference in (2.4) and(3.2) (in frontof h2 and λ2) shouldbe interpretedwith care in the twomodels. This differencemay have significant impact on the physical quantities of interest. For instance, in the slow oscillatory flow of a viscous fluid past a sphere (Smith, 1995), the vorticity on the sphere (r =1) is given by ωt = [3 2 (σ+1) ] eαt (4.5) where σ2 = ih. For the slow flow through porous media, the vorticity on the sphere (Pop and Ingham, 1996) is ωB = [3 2 (σ+1) ] (4.6) with σ = λ. The vorticity given in (4.5) vanishes for some values of σ (negative values of σ are admissible in this case) implying that flow separation is possible. However, the vorticity given in (4.6) never vanishes for σ > 0 (σ cannot take negative values here) and consequently there is no flow separation for this flow in the Brinkmanmedium. On some general solutions of transient Stokes and Brinkman equations 413 5. Conclusion General solutions for the transient Stokes flow and flow through porous media are discussed in this paper. Differential representation for transient flow is shown to be equivalent to Lamb’s general solution and new formulae connecting the two solutions are given. An alternative repre- sentation for bounded flows constrained by a plane wall is also provided. The usefulness of this new solution representation for flows involving disks is briefly outlined. It is observed that the solution representations are geometry dependent. The analogy between the oscillatory flows and the flow through porous media is exploited to derive the solutions of the Brinkman equations. Some representative boundary value problems are considered to justify our observations concer- ning the two models. Our discussion indicates that the solutions of Brinkman models can be derived from the transient Stoke flowmodels and so the duplicationmaybe avoided. Apart from genuinemathematical interest, the results provided heremay found useful in practice where the general and generalized solutions (Shu and Chwang, 2001) are needed to understand the basic engineering aspects in fluid and elastic environments. Time-dependentmultiparticle interactions andmobility of particles close to theBrinkman half-space (Damiano et al., 2004) including elec- trophoreticmotion of charged particles in porousmedia (Tsai et al., 2011) are a few prospective topics for further research using our general solution representations. References 1. Basset A.B., 1888,A Treatise on Hydrodynamics, vol. 3, Ch. 21, 22, Cambridge, Deighton Bell 2. Bhatt B.S., Sacheti N.C., 1994, Flow past a porous spherical shell using the Brinkmanmodel, Journal of Physics D: Applied Physics, 27, 37-41 3. Boussinesq M.J., 1885,Application des Potentiels a l’Étude de l’Équilibre et duMouvements des Solides Élastiques, Gaythier-Villars, Paris 4. BrinkmanH.C., 1947,A calculation of the viscous force exerted by a flowing fluid dense swarmof particles,Applied Science Research,A1, 27-34; On the permeability of media consisting of closely packed porous particles,Applied Science Research,A1, 81-86 5. Damiano E., Long R.D.S., El-Khatib F.H., Stace T.M., 2004,On themotion of a sphere in a Stokes flow parallel to a Brinkman half-space, Journal of Fluid Mechanics, 500, 75-101 6. Dassios G., Vafeas P., 2004, Comparison of differential representations for radially symmetric Stokes flow,Abstract and Applied Analysis, 2004:4, 347-360 7. Felderhof B.U., 1978, Force density induced on a sphere in linear hydrodynamics: I. Fixed sphere, stick boundary conditions,Physica A, 84, 557-568 8. Feng J., Ganatos P., Weinbaum S., 1998, The general motion of a circular disk in Brinkman medium, Physics of Fluids, 10, 2137-2146 9. Happel J., Brenner H., 1983, Low Reynolds Number Hydrodynamics, The Hague: Martinus Nijhoff 10. Higdon J.J.L., KojimaM., 1981,On the calculation od Stokes flowpast porous particles, Inter- national Journal of Multiphase Flow, 7, 719-728 11. Howells I.D., 1974, Drag due to motion of a Newtonian fluid through a sparse random array of fixed rigid objects, Journal of Fluid Mechanics, 64, 449-475 12. Kim S., Russel W.B., 1985, The hydrodynamic interactions between two spheres in Brinkman medium, Journal of Fluid Mechanics, 154, 253-268 13. Lamb H., 1945,Hydrodynamics, 6th edition, Cambridge University Press, Cambridge 1932; New York, Dover 414 D. Palaniappan 14. Lawrence C.J., Weinbaum S., 1986, The force on an axisymmetric body in linearized time- dependent motion: a newmemory term, Journal of Fluid Mechanics, 171, 209-218 15. Lundgren T.S., 1972, Slow flow through stationary random beds and suspension of spheres, Journal of Fluid Mechanics, 51, 273-299 16. Mazur P., Bedeaux D., 1974, A generalization of Faxen’s theorem to non-steady motion of a sphere through an incompressible fluid in arbitrary flow,Physica A, 76, 235 17. Masliyah J.H., Neale G., Malysa K., Van de ven T.G.M., 1987, Creeping flow over a composite sphere: Solid core with porous shell,Chemical Engineering Science, 42, 245-253 18. Milne-Thomson L.M., 1968,Theoretical Hydrodynamics, Macmillan, London 19. NaghdiP.M.,HsuC.S., 1961,On the representationof displacements in linear elasticity in terms of three stress functions, Journal of Mathematics and Mechanics, 10, 233-245 20. NealeG., Epstein N., NaderW., 1973, Creeping flow relative to permeable spheres,Chemical Engineering Science, 28, 1865-1874 21. Neuber N., 1934, Ein Neuer Ansatz Zur Lösung räumblicher Probleme der Elastizitätstheorie, Zeitschrift für Angewandte Mathematik und Mechanik, 14, 203-212 22. Nield D.A., Bejan A., 1992,Convection in Porous Media, Springer-Verlag 23. Oomes G., Mijulieff P.E., Beckers H.L., 1970, Frictional force exerted by a flowing fluid on a permeable particle with particular reference to polymer coils, Journal of Chemical Physics, 53, 4123-4130 24. Padmavathi B. S., Amaranath T., 1996, Stokes flow past a composite porous spherical shell with a solid core,Archives of Mechanics, 48, 311-323 25. PadmavathiB.S., SekharG.P.R.,AmaranathT., 1998,Anote on complete general solutions of Stokes equations,Quarterly Journal of Mechanics and Applied Mathematics, 51, 383-388 26. Palaniappan D., 2000, General slow viscous flows in a two-fluid system, Acta Mechanica, 139, 1/4, 1-13 27. Palaniappan D., 2009, Some general differential representations and theorems involving bounda- ries in multiphase Stokes flow, Theoretical Methods for Micro Scale Viscous Flows, eds. Francois Feuillebois and Antoine Sellier, 19-36 28. Palaniappan D., Nigam S.D., Amaranath T., Usha R., 1992, Lamb’s solution of Stokes equations: A sphere theorems,Quarterly Journal of Mechanics and Applied Mathematics, 45, 47- 56 29. Pop I., Ingham D.B., 1996, Flow past a sphere embedded in a porous medium based on the Brinkmanmodel, International Communications in Heat and Mass Transfer, 23, 865-874 30. PozrikidisC., 1991,SingularityMethods for LinearisedViscousFlow,CambridgeUniversityPress 31. Pozrikidis C., 1994, A bibliographical note on the unsteady motion of a spherical drop at low Reynolds number,Physics of Fluids, 6, 9, 3209 32. QinY.,KaloniP.N., 1993,Creepingflowpastaporousspherical shell,Zeitschrift fürAngewandte Mathematik und Mechanik, 73, 77-84 33. Saffman P.G., 1971, On the boundary conditions on the surface of a porousmedium, Studies in Applied Mathematics, 50, 93-101 34. Shu J.-J., Chwang A.T., 2001, Generalized fundamental solutions for unsteady viscous flows, Physical Review E, 63, 051201 35. Smith S. H., 1995, Structural changes in transient Stokes flow, Quarterly Journal of Mechanics and Applied Mathematics, 48, 285-309 36. Stokes G.G., 1851, On the effect of fluids on the motion of pendulums, Transactions of the Cambridge Philosophical Society, 9, 8 On some general solutions of transient Stokes and Brinkman equations 415 37. Tam C.K.W., 1969, The drag on a cloud of spherical particles in low-Reynolds number flow, Journal of Fluid Mechanics, 38, 537-546 38. Tsai P., Huang C.H., Lee E., 2011, Electrophoresis of a charged colloidal particle in porous media: boundary effect of a solid plane, Langmuir, 27, 22, 13481-13488 39. Watson G.N., 1952,A Treatise on the Theory of Bessel Functions, 2nd edition, Cambridge Uni- versity Press 40. Yang S.-M., Leal L.G., 1991, A note on memory-integral contributions to the force on an accelerating spherical drop at lowReynolds number,Physics of Fluids, A3, 7, 1822-1824 41. Zhang W., Stone H.A., 1998, Oscillatory motions of circular disks and near spherical particles in viscous flows, Journal of Fluid Mechanics, 367, 329-348 Manuscript received June 30, 2013; accepted for print October 28, 2013