Abstract
The aim of this article is to advance the current state of knowledge for steady, isothermal, incompressible, laminar flow within a channel featuring a non-zero tangential (or slip) velocity at the permeable walls. There has been significant interest in understanding the solutions to these problems. However, a firm mathematical understanding of the solutions to the slip problem and their properties is yet to be fully developed. For example, we still do not know: if the slip problem is well-posed; where the precise solution lies; if and how approximations converge to the solution; and what the estimates on approximation errors are. Herein we formulate a new mathematical foundation that includes existence; uniqueness; location; approximation; convergence and error estimates. Our strategy involves developing insight via new and interesting connections between the boundary value problem arising from modelling the laminar flow with slip velocity, and the theory of fixed points of operators.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
Porous flow problems are found in many physical phenomena, including hydraulic engineering (Chellam and Wiesner 1993), ultrafiltration membranes (Singh and Laurence 1979), gasesous diffusion (Wang et al., 1990), biological organisms (Majdalani et al. 2002) and ablation cooling (Dauenhauer and Majdalani 2003), and thus the challenge of understanding solutions to these problems have continued to interest the scientific research community.
A seminal study on porous flow problems is due to Berman (1953), who analyzed the velocity field of a homogeneous fluid that was laminarly flowing in a channel formed by two porous walls. Therein, Berman assumed a non-slip (or zero tangential velocity) boundary condition at the porous walls, which characterizes flows with solid bounding walls (Singh and Laurence 1979). However, the work of several authors (Beavers and Joseph 1967; Kohler 1973; Saffman 1971) suggests that laminar flow problems with porous surfaces do exhibit a slip boundary condition. For example, when the fluid contains smaller particles that pass through the porous wall material, and larger particles that accumulate near the porous boundaries, it creates a so-called “slip" boundary effect. For instance, the physical phenomena of concentration polarization (Blatt et al. 1970 [p. 47.]; Michaels 1968 [p.297]), involves anisotropic ultrafiltration membranes and an accumulation of solutes that forms “a polarized gel-layer” (Singh and Laurence 1979 [p.721]). As such, flow problems with slip conditions have been examined subsequently by many authors, such as Chellam et al. (1993, 1992); Ullah et al. (2021); Bhat and Katagi (2021); Balhoff et al. (2010); Ullah et al. (2020); Siddiqui et al. (2020); Varunkumar and Muthu (2020); Ullah et al. (2019); Farooq et al. (2018); Ashwini et al. (2017); Cox and Hill (2011); El-Genk and Yang (2009); Nazari Moghaddam and Jamiolahmady (2016); Nishiyama and Yokoyama (2017); Sparrow et al. (1974); Rasoulzadeh and Panfilov (2018); Skjetne and Auriault (1999); Taherinejad et al. (2021).
We analyze the following nonlinear, ordinary differential equation
with \(f= f(\eta )\) a part of a stream function associated with the components of the fluid’s velocity, \({\mathcal {R}}\) is a Reynolds number and [0, 1] covers half of the width of the channel (a more detailed explanation will be given in Sect. 2). The differential equation (1.1) is coupled with the boundary conditions:
where \(\gamma\) is a non-negative constant that we refer to as the slip coefficient. If \(\gamma = 0\) then (1.1), (1.2) collectively reduce to the classical non-slip case of Berman (1953).
Several scholars (Singh and Laurence 1979; Chellam et al. 1992; Chellam and Wiesner 1993; Chellam et al. 1993; Chellam and Lui 2006; Guo et al. 2020) have developed approximations to solutions to porous flow problems that include type (1.1), (1.2) with suction (\(\mathcal{R} >0\)) through perturbation techniques, numerical integration and Runge-Kutta schemes. For instance, a polynomial approximation is generated by constructing a perturbation expansion, with standard practice to usually consider two terms or less. The expansions are done in terms of a parameter, in this case \({{{\mathcal {R}}}}\), the Reynolds number. Essentially, this involves starting with a solution for \({{{\mathcal {R}}}} = 0\) and then constructing approximative (but low degree) polynomials, see, for example, Singh and Laurence (1979). The problem (1.1), (1.2) has also been transformed via a substitution so that the Reynolds numbers was transferred to one of the boundary conditions (Chellam and Wiesner 1993) and then Runge-Kutta methods were used to approximate solutions. Shooting methods via numerical integration have also been applied to develop insight into the solutions of (1.1), (1.2), for example, in Chellam and Lui (2006).
However, a firmer and more precise mathematical understanding of the solutions to porous flow problems with slip conditions and their properties is yet to be fully developed. For example, it is unknown if (1.1), (1.2) is well-posed for the slip case \(\gamma > 0\), the location of the solution(s) remains unknown, it is yet to be known if and how approximations converge to this solution, and what the estimates for errors in the approximations are. Herein we formulate a new mathematical foundation regarding solutions, including existence; uniqueness; location; approximation; convergence and error estimates. Our strategy involves developing new and interesting connections between solutions to (1.1), (1.2) and the theory of fixed points of operators, and cover both suction cases \(({{{\mathcal {R}}}}>0)\) and injection cases \((\mathcal{R}< 0)\).
A real valued function on [0, 1] with a continuous fourth order derivative is called a solution to the BVP (1.1), (1.2) if it satisfies both (1.1) and (1.2) for some value of \({\mathcal {R}}\) and some value of \(\gamma \ge 0\).
Recently, Almuthaybiri and Tisdell (2022) analyzed the non-slip case (\(\gamma = 0\)) of (1.1), (1.2) and employed a contractive mapping technique to better understand the nature of solutions. They established existence, uniqueness and approximation of solutions under the assumptions
Their work naturally raises questions regarding solutions to the more accurate slip case \(\gamma > 0\), such as: for what values of \({{{\mathcal {R}}}}\) and \(\gamma >0\) does (1.1), (1.2) have a unique solution? Where is this solution located? How can we approximate this solution? We address these questions herein through the use of fixed point methods, in particular, via contractive mappings. We discover that for all slip coefficients \(\gamma >0\), the BVP (1.1), (1.2) has a unique solution for sufficiently small values of \(|{{{\mathcal {R}}}}|\), and we determine a location for this solution. Furthermore, we construct a sequence of approximating functions whose limit is the above solution to (1.1), (1.2). The convergence rate is shown to be linear, and we establish error estimates on the approximative sequence, which generates approximations to any prescribed level of accuracy. We also compare our results with that of the non-slip problem to shed light on connections and differences between the two.
The layout of this article is as follows. In Sect. 2 we briefly derive the problem (1.1), (1.2) to introduce notation and to provide context. We also formulate an equivalent integral equation to the BVP (1.1), (1.2). In Sect. 3 we construct new bounds on the integrals of various Green’s functions for (1.1), (1.2) and relate them to the non-slip case. In Sect. 4 we synthesize the equivalent integral representation and our bounds to enable an application of fixed point theory to a suitable operator that yields new existence, uniqueness and approximation results. In Sect. 5 we discuss an open problem.
2 Problem Derivation and an Integral Form
Let us briefly derive the problem (1.1), (1.2) under consideration. For additional details see, for example, Singh and Laurence (1979) or Chellam and Wiesner (1993).
Consider steady, laminar flow of fluid containing macro-molecular solutes in a two-dimensional channel with porous walls in the form of two parallel and flat ultrafiltration membranes, where the walls are separated by a distance of 2h, see Fig. 1. We assume the fluid has constant density \(\rho\) and kinematic viscosity \(\nu\). The fluid is subject to either injection or suction with constant velocity \({\mathcal {V}}\) through the walls, with each wall having equal permeability. This implies the flow will be symmetrical about the midplane of the channel.
Choose a coordinate system with its origin located at the center of the channel. We let x denote the co-ordinate axis that is parallel to the channel walls, and let y be the axis that is perpendicular to the channel walls. Also, let \(u=u(x,y)\) represent the velocity component of the fluid in the x direction and let \(v=v(x,y)\) represent the velocity component of the fluid in the y direction. Let p denote the pressure.
Introduce the normalized variable
and form the Navier-Stokes equations
The continuity equation is
and the associated boundary conditions are
Note that the slip boundary condition captures the assumption that the slip velocity at the porous boundaries is proportional to the wall shear rate (Chellam and Wiesner 1993). Here \(k \ge 0\) is the permeability of the membrane matrix and \(\alpha >0\) is a dimensionless constant that depends on the surface characteristics of the membrane (Singh and Laurence 1979). Throughout this paper, the constant \(\gamma\) will represent the slip coefficient associated with the slip condition at the boundaries, namely
Due to symmetry, we can focus our attention on (say) the upper half-channel of thickness h. Since our attention is on the upper half channel from \(\eta = 0\) to \(\eta = 1\), we disregard the boundary conditions involving \(u(x,-1)\) and \(v(x,-1)\) from what follows.
A stream function \(\psi\) exists with
with the continuity equation holding.
Drawing on Berman’s assumption (Berman , 1953) that the velocity component v is independent of x, we can introduce a stream function, \(\psi\), that takes the form
Above, f is a function of \(\eta\) which is to be determined later, and \({{{\bar{u}}}}(0)\) is an arbitrary velocity at \(x = 0\) that is managed away.
From (2.1) we can derive the velocity components in terms of f
Our u and v in (2.2) are expressed in terms of f and \(f'\), with the equations of motion then leading to (1.1) with \({{{\mathcal {R}}}} = {\mathcal {V}} h / \nu\) a Reynolds number. Furthermore, if we consider (2.2) at the boundaries of the half-channel bounded by \(\eta = 0\) and \(\eta = 1\) then we obtain (1.2).
Let us establish a new result that links the equivalency between the slip BVP (1.1), (1.2) and an integral equation. The integral equation will be drawn upon in Sect. 3 to leverage our main results.
Theorem 2.1
For all \(\gamma \ge 0\), the BVP (1.1), (1.2) is equivalent to the integral equation
Above, \(G(\eta ,s)\) is the Green’s function
and \(\phi\) is given by
Proof
The proof is motivated by that of Almuthaybiri and Tisdell (2022) for the non-slip case and we make the appropriate modifications to incorporate the slip conditions (1.2). Consider the form
where \(\phi\) is a solution to
and \(\phi _1\) is a solution to
Direct differentiation of f and substitutions into (1.1), (1.2) shows that our f satisfies these equations and thus is of a suitable form.
The general solution to the homogeneous differential equation in (2.6) can be constructed via integration, forming a general polynomial of degree three, namely
The associated boundary conditions in (2.6) are then applied to determine the coefficients. For example, \(\phi (0)=0\) yields \(D_0=0\), and \(\phi ''(0) = 0\) yields \(B_0=0\). Furthermore, the remaining boundary conditions produce
which are solved for \(A_0 = -1/[2(1 + 3\gamma )]\) and \(C_0= 3(1+2\gamma )/[2(1 + 3\gamma )]\) to obtain
Consider the nonhomogeneous differential equation for \(\phi _1\) in (2.7). We integrate both sides from \(s = 0\) to \(s=\eta\) four times to produce
where A, B, C, D are arbitrary constants of integration which we determine from the boundary conditions in (2.7). The boundary data \(\phi _1(0)=0\) leads to \(D=0\) and \(\phi _1''(0)=0\) produces \(B=0\). If we now draw on the right-hand boundary data, then we obtain
The above simultaneous equations are solved for A and C to obtain
Substituting these expressions into (2.8) and collecting like terms leads us to the form (2.3).
If we differentiate our f with the above values of A and C, then we obtain the differential equation (1.1). If we evaluate our integral representation for f and its derivatives at the boundary points, then we see that the boundary conditions (1.2) also hold. \(\square\)
Remark 2.1
Our derived formulae for the Green’s function G in (2.4) and \(\phi\) in Theorem 2.1 extends the result of Almuthaybiri and Tisdell (2022) from \(\gamma = 0\) to \(\gamma \ge 0\).
3 Bounds on the Green’s Functions
In this section we develop new bounds on the integral involving the absolute value of the Green’s function in (2.4) and its derivatives. These bounds will be applied in Sect. 4 to help generate our theorems on existence, uniqueness and approximation of solutions to (1.1), (1.2).
Before we calculate new bounds, we notice an important connection between our Green’s function (2.4) for the slip case, and the Green’s function for the non-slip problem derived by Almuthaybiri and Tisdell (2022). For all \((\eta ,s) \in [0,1] \times [0,1]\) and \(\gamma \ge 0\), we have
with
and
Here, \(H(\eta ,s)\) is the Green’s function for the non-slip problem when \(\gamma = 0\) that was derived by Almuthaybiri and Tisdell (2022), and \(F(\eta ,s)\) is an additional component that arises from the slip condition.
If we apply the triangle inequality to (3.1) then for all \(\gamma \ge 0\) we can deduce the inequality
Similarly, for the partial derivatives, we have
Integrating both sides of (3.4) from 0 to 1 gives, for all \(\eta \in [0,1]\) and \(\gamma \ge 0\),
Using the above connections to the non-slip problem, our strategy is to analyze the integrals involving |F| and |H| separately, and to synthesize them with Almuthaybiri and Tisdell (2022) results to produce new bounds. This strategy streamlines some of the calculations and essentially reduces the problem of constructing bounds to only working with F, which has nice symmetric properties and is relatively easy to differentiate, integrate and analyze. In doing so, we hope to lead the reader to the heart of the problem without experiencing significant departures. Our method of analyzing integrals involving |F| and |H| separately means that our bounds may not be the sharpest possible.
Our new bounds are formulated in terms of the slip coefficient \(\gamma\), and then independently of \(\gamma\). Including \(\gamma\) in our estimates illustrates the connection between the slip and non-slip cases, while excluding \(\gamma\) enables the bounds to be of less complex nature, and perhaps more user-friendly.
Theorem 3.1
For all \(\gamma \ge 0\), the Green’s function G in (2.4) satisfies \(G\le 0\) on \([0,1]\times [0,1]\) and
Proof
From Almuthaybiri and Tisdell (2022), we have our H in (3.2) satisfying \(H\le 0\) on \([0,1]\times [0,1]\) and
Now, we claim that for F in (3.3), we have \(F\le 0\) on \([0,1]\times [0,1]\). If we consider the case \(0 \le s \le \eta \le 1\) then we have \(1-\eta \ge 0\) and \(s\ge 0\), and
Hence,
Employing a similar technique, but interchanging s and \(\eta\), we have
Combining the above two cases, we conclude that \(F\le 0\) on \([0,1]\times [0,1]\). Thus, the form (3.1) gives us \(G \le 0\) on \([0,1]\times [0,1]\) as claimed.
Consequently, for all \(\eta \in [0,1]\) we have
And for all \(\eta \in [0,1]\),
This quadric function achieves its minimum of \(-5/64\) on [0, 1] at \(\eta = 1/2\), and so for all \(\eta \in [0,1]\) we have
Thus, (3.7) leads to
\(\square\)
Let us continue the momentum of the previous theorem, but now focusing on \(|\partial G/\partial \eta |\).
Theorem 3.2
For all \(\gamma \ge 0\), the Green’s function G in (2.4) satisfies
Proof
Let us work with (3.5) for the case \(i=1\). By Almuthaybiri and Tisdell (2022), we have
Thus, our attention turns to establishing an estimate on \(\int _0^1 |\partial F/\partial \eta | \ {\rm d}s\). Differentiating F in (3.3) with respect to \(\eta\) gives:
Thus, for all \(\eta \in [0,1]\) we have
This quadric function achieves its maximum value on [0, 1] of 415/324 at \(\eta = 2/3\) and so
Substituting the estimates (3.9) and (3.10) into (3.5) with \(i=1\) establishes our estimate (3.8). \(\square\)
We now develop bounds involving higher order derivatives of G as follows.
Theorem 3.3
For all \(\gamma \ge 0\), the Green’s function G in (2.4) satisfies
Proof
From Almuthaybiri and Tisdell (2022), we have
Thus, our challenge is to construct a bound on the remaining integral in (3.5) when \(i=2\). Differentiating F in (3.3) with respect to \(\eta\) twice gives:
Clearly we have \(\dfrac{\partial ^2}{\partial \eta ^2}F(\eta ,s) \ge 0\) for all \((\eta ,s) \in [0,1] \times [0,1]\) and so
This above quadratic function achieves its maximum of 3/4 on [0, 1] at \(\eta = 1/2\) and thus for all \(\eta \in [0,1]\) we have
If we now leverage the above two bounds then (3.5) when \(i=2\) yields (3.11). \(\square\)
Our final bound result is as follows.
Theorem 3.4
For all \(\gamma \ge 0\), the Green’s function G in (2.4) satisfies
Proof
From Almuthaybiri and Tisdell (2022), we know that
Differentiating F with respect to \(\eta\) three times gives:
It clear that \(\dfrac{\partial ^3}{\partial \eta ^3}F(\eta ,s) \le 0\) for \(0 \le s \le \eta \le 1\) and \(\dfrac{\partial ^3}{\partial \eta ^3}F(\eta ,s) \ge 0\) for \(0 \le \eta \le s \le 1\), and
This quadratic function achieves its maximum at the boundaries \(\eta = 0\) and \(\eta = 1\), so that for all \(\eta \in [0,1]\) we have
Substituting our two bounds into (3.5) when \(i=3\) yields (3.12). \(\square\)
4 Existence, Uniqueness and Approximation
This section contains our main theorems regarding the existence, uniqueness, location and approximation of solutions.
4.1 Complete Metric Space, Bound and Lipschitz Constants
We will work in the context of a complete metric space. Let \(C^3([0,1])\) denote the set of real-valued functions that are defined on [0, 1] and have a continuous third order derivative. Consider the following metric on \(C^3([0,1])\):
where
and each \(\beta _i\) is defined in (3.6), (3.8), (3.11) and (3.12). The metric space \((C^3([0,1]),d)\) is known to be complete (see (Almuthaybiri and Tisdell 2022)).
Let \(R>0\) be a constant and let \(\phi\) be defined in (2.5). In our main results, it will be shown that the following set provides a suitable location for the graph of the unique solution to (1.1), (1.2)
We note that for all \(\gamma \ge 0\), our \(\phi\) in (2.5) satisfies
To enable the invariance of an appropriate operator between sets, the following result will be helpful by establishing a bound on (1.1) on B.
Theorem 4.1
Let
We claim that for all \(\gamma \ge 0\) our |h| is bounded on B by
Proof
The proof is similar to that of Almuthaybiri and Tisdell (2022) with appropriate modifications made to accommodate the more general bounds from (4.1). For \((\eta ,u,v,w,z) \in B\) consider
\(\square\)
The following result constructs new Lipschitz constants for (1.1) on B for the non-slip case.
Theorem 4.2
For given \(R>0\), \({{{\mathcal {R}}}}\) and \(\gamma \ge 0\), the function
satisfies
for some constants \(L_i\).
Proof
Here we take a slightly different approach than Almuthaybiri and Tisdell (2022) who produced different Lipschitz constants for the case \(\gamma = 0\). We opt for algebraic techniques, rather than constructing bounds on the partial derivatives of h on B. One important reason for including some of the details of this proof is to illustrate the precise form of the Lipschitz constants \(L_i\).
For all \((\eta ,u_0,u_1,u_2,u_3), (\eta ,v_0,v_1,v_2,v_3) \in B\) consider
Now, on B,
Also, we can continue this algebraic process for each of the remaining terms in (4.2) to obtain
\(\square\)
4.2 Contraction Mapping Approach
We will employ the following well known fixed point theorem found in Zeidler (1986)[Theorem 1.A] to generate a unique fixed point of an operator connected with (1.1), (1.2).
Theorem 4.3
Let X be a nonempty set and let d be a metric on X such that (X, d) forms a complete metric space. If the mapping \(T:X \rightarrow X\) satisfies
then there is a unique \(z\in X\) such that \(Tz = z\). In addition, for any \(z_0\in X\) we have \(d(z_n,z) \rightarrow 0\) where \(z_n\) is a recursively defined sequence defined via \(z_{n+1} :=Tz_n\).
Armed with the integral equation of Sect. 2, the bounds of Sect. 3, and the Lipschitz constants and bound of Sect. 4, we are now ready to formulate our main theorems.
Theorem 4.4
If there is a \(R>0\) and \({\mathcal {R}}\) such that
then for all \(\gamma \ge 0\) the slip BVP (1.1), (1.2) has a unique solution f such that
Proof
The structure of the proof follows similar lines as that of Almuthaybiri and Tisdell (2022) and so is only sketched, highlighting the distinctions.
Choose a suitable value of \(R>0\) to form B, with R and \({\mathcal {R}}\) satisfying (4.4) and (4.5).
Consider the set
with \(({{{\mathcal {B}}}}_R,d)\) forming a complete metric space.
Consider the operator \(T:{{{\mathcal {B}}}}_R \rightarrow C^3([0,1])\) that we define by
Let us show that T has a unique fixed point in \({{{\mathcal {B}}}}_R\), by illustrating that the assumptions of Theorem 4.3 hold with \(X={{{\mathcal {B}}}}_R\).
Firstly, we claim \(T:{{{\mathcal {B}}}}_R \rightarrow {{{\mathcal {B}}}}_R\). For \(f\in {{{\mathcal {B}}}}_R\) and \(\eta \in [0,1]\), we have
Furthermore, we have
and so
A similar line of arguments also give
so that
Combining the above inequalities, for all \(f\in {{{\mathcal {B}}}}_R\) we have
where we have invoked (4.4). Thus, for all \(f\in {{{\mathcal {B}}}}_R\) we see that \(Tf \in {{{\mathcal {B}}}}_R\) and hence \(T:{{{\mathcal {B}}}}_R \rightarrow {{{\mathcal {B}}}}_R\).
Secondly, we claim that T is contractive on \({\mathcal {B}}_R\) in the sense of (4.3). For \(f,g \in {\mathcal {B}}_R\) and \(\eta \in [0,1]\), consider
where we invoked the bound from Theorem 4.2.
In a similar fashion, we can also show
Thus, for all \(f,g \in {\mathcal {B}}_R\) we have
Due to our assumption (4.5) we see that T is a contractive map on \({\mathcal {B}}_R\).
Thus, the conditions of Theorem 4.3 are satisfied for our T and \(X={\mathcal {B}}_R\). We conclude that T has a unique fixed point in \({\mathcal {B}}_R\subset C^3([0,1])\). This is equivalent to proving that our slip BVP (1.1), (1.2) has a unique solution for all \(\gamma \ge 0\). \(\square\)
Let us delve deeper into our assumptions (4.4) and (4.5) by establishing some more concrete values for \({{{\mathcal {R}}}}\), R and \(\gamma\) that ensure (4.4) and (4.5) hold.
Theorem 4.5
For all \(\gamma \ge 0\) and
the slip BVP (1.1), (1.2) admits a unique solution whose graph lies completely in B with
Proof
We see that (4.4) and (4.5) can be rewritten as
The crossing of the graphs of the right hand sides in (4.7), (4.8) occurs when
The “value” of \(|{{{\mathcal {R}}}}|\) at this point is
and so, for values of \(|{\mathcal {R}}|\) strictly less than (4.9), both of our inequalities (4.4) and (4.5) will hold. Thus, for the above R and range of \({\mathcal {R}}\) the conclusion of Theorem 4.4 holds. \(\square\)
Remark 4.1
If we compare the bounds on \({{{\mathcal {R}}}}\) in (4.6) that includes the slip case \(\gamma >0\), with the bounds on \({{{\mathcal {R}}}}\) for the non-slip case of Almuthaybiri and Tisdell (2022) in (1.3), then we observe that the range of \(\mathcal{R}\) values in the non-slip BVP is about half of that in the slip BVP. However, in the slip case we have results for all \(\gamma > 0\) that also include the special case \(\gamma = 0\). So, to modify an old saying about swings and roundabouts, when comparing our results of the slip situation with that of the non-slip situation, it appears to be a case of: what you lose on the Reynolds number, you gain on the slip constant.
Theorem 4.5 also ensures insights into the approximation of solutions to the slip BVP (1.1), (1.2).
Theorem 4.6
Let the conditions of Theorem 4.5 hold. If we define a sequence of functions \(f_n = f_n(\eta )\) on [0, 1] via
then for all \(\gamma \ge 0\), our \(f_n\) converges to the solution f of (1.1), (1.2) lying in B, and the rate of convergence is given by
In addition, for each n, an a priori estimate on the error is given by
and for each n, an a posteriori estimate on the error is given by
Proof
The style of proof is well known and follows from the conditions of Theorem 4.3 holding. For example, the (linear) rate of convergence can be shown in the following way. Since T satisfies \(Tf = f\) and \(f_{n+1} = Tf_n\) we have
where we have used the contraction property of T with
Furthermore, the a priori error on the estimate can be shown via repeatedly applying the triangle inequality and the contraction property of T:
For brevity, we omit the proof of the a posteriori estimate on the error. \(\square\)
Remark 4.2
In fact, it is not necessary to start with the particular form of \(\phi _0\) in Theorem 4.6. We can begin our recursive sequence of approximations with any function \(f_0\in C^3([0,1])\) such that
and we will still have \(f_n\) converging on [0, 1] to the solution f in the sense of Theorem 4.6.
Let us briefly compare the approximation method using our integral form in Theorem 4.6 with that of the perturbation approach in Singh and Laurence (1979). Therein, the following equivalent form of (1.1) is considered
with (1.2), where K is a constant of integration. Singh and Laurence (1979) sought a solution for small \({{{\mathcal {R}}}}\) in series form
where
and each \(g_i\) and \(K_i\) are assumed to be independent of \(\mathcal{R}\). They collected corresponding coefficients for powers of \({\mathcal {R}}\) and constructed \(g_0\) and \(g_1\). There appear to be limitations in this particular approach, in the sense that there are a significant number of constants to compute at each stage of the process, and the boundary conditions are drawn on at every iteration for each \(g_i\). For instance, keeping the constant K in the differential equation under consideration and expanding it as a power series means that there are many \(K_i\) to be found. In addition, for the zero order, \(g_0''' = K_0\) is solved easily, however there are now four separate boundary conditions to navigate. For the first order, \(g_1''' = K_1 - (g_0')^2 + g_0 g_0''\), there are again four separate boundary conditions to incorporate, and so on in each iteration. If we compare the perturbation approach with our scheme in Theorem 4.6, then we see that our scheme of approximations does not include the constant K or rely on repeated incorporation of boundary conditions at every step of the iteration process. Thus, we argue that our strategic use of an integral representation throughout this article forms a more streamlined approach than has been previously available for the slip problem (1.1), (1.2).
5 Opportunities and Conclusion
We note that the estimates used from Sect. 2 in Sects. 3 and 4 are independent of the slip coefficient \(\gamma\). This strategy has been done for simplicity, but raises the open question of what would happen if the results of Sects. 3 and 4 were formulated with the \(\gamma\)-dependent bounds of Section 2. We suspect that this would lead to extremely complex expressions involving \(\gamma\), but might improve the range of Reynolds numbers set in this paper, with the interval tending to that of Almuthaybiri and Tisdell (2022) when \(\gamma \rightarrow 0\).
We also note that the existence, uniqueness and approximation results of this paper apply to the bounded set B that contains the graph of the solution. The raises questions regarding what happens outside of the set B. Do more solutions exist there? How many, and what are their properties? This naturally aligns with the position of Kuo and Wang (2012) that a precise understanding of the multiplicity of solutions appears to be still open. As a very simple motivational example, consider the equation \(u = 1 + {{{\mathcal {R}}}} u^2\). For \(X = \{u \in R: |u-1| \le 1\}\) and \(T(u) = 1 + {{{\mathcal {R}}}}u^2\), Theorem 4.3 can be applied to show the existence and uniqueness of a solution to our equation for \(|{{{\mathcal {R}}}}| < 1/4\). However, for \(0<{{{\mathcal {R}}}} < 1/4\) there is a second solution \(u^*\) satisfying \(u^* > 2\). Thus, additional solutions can exist that do not lie in the original set under consideration.
The main contribution of the ideas in this paper involved constructing new understanding of solutions to the slip BVP (1.1), (1.2). We now have novel knowledge regarding the existence, uniqueness, location and approximation of the precise solution, including a specific range on the Reynolds number that guarantees this, for all slip coefficients.
Data Availability Statement
Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.
Change history
25 February 2023
Missing Open Access funding information has been added in the Funding Note.
References
Almuthaybiri, Saleh S., Tisdell, Christopher C.: Laminar flow in channels with porous walls: advancing the existence, uniqueness and approximation of solutions via fixed point approaches. J. Fixed Point Anal. Appl. 24, 55 (2022). https://doi.org/10.1007/s11784-022-00971-8
Ashwini, B., Katagi, N.N., Rai, A.S.: Analysis of laminar flow through a porous channel with velocity slip. Malays. J. Math. Sci. 11(3), 423–439 (2017)
Balhoff, M., Mikelić, A., Wheeler, M.F.: Polynomial filtration laws for low reynolds number flows through porous media. Transp. Porous Med. 81, 35 (2010). https://doi.org/10.1007/s11242-009-9388-z
Beavers, G.S., Joseph, D.D.: Boundary conditions at a naturally permeable wall. J. Fluid Mech. 30, 197–207 (1967)
Berman, A.S.: Laminar flow in channels with porous walls. J. Appl. Phys. 24(9), 1232–1235 (1953). https://doi.org/10.1063/1.1721476
Bhat, A., Katagi, N.N.: Keller-box solution of the stagnation point micropolar fluid flow between porous plates with injection. Adv. Math. Sci. J. 10(1), 517–526 (2021). https://doi.org/10.37418/amsj.10.1.51
Blatt, W.F., Dravid, A., Michaels, A.S., Nelsen, L.: Solute polarization and cake formation in membrane ultrafiltration: causes, consequences, and control techniques. In: Flinn, J.E. (ed.) Membrane Science and Technology. Plenum Press, New York (1970)
Chellam, S., Lui, M.: Effect of slip on existence, uniqueness, and behavior of similarity solutions for steady incompressible laminar flow in porous tubes and channels. Phys. Fluids 18, 083601 (2006). https://doi.org/10.1063/1.2236302
Chellam, S., Wiesner, M.R., Dawson, C.: Slip at a uniformly porous boundary: effect on fluid flow and mass transfer. J. Eng. Math. 26(4), 481–492 (1992). https://doi.org/10.1007/bf00042765
Chellam, S, Wiesner, M.R.: Laminar flow with slip in channels with uniformly porous walls. J. Hydraul. Eng. 119(1), 126–132 (1993). https://doi.org/10.1061/(ASCE)0733-9429(1993)119:1(126)
Chellam, S., Wiesner, Mark R.: Slip flow through porous media with permeable boundaries: implications for the dimensional scaling of packed beds. Water Environ. Res. 65(6), 744–749 (1993). https://www.jstor.org/stable/25044372
Cox, B., Hill, J.: Flow through a circular tube with a permeable Navier slip boundary. Nanoscale Res. Lett. 6(1), 389 (2011). https://doi.org/10.1186/1556-276X-6-389
Dauenhauer, Eric C., Majdalani, Joseph: Exact self-similarity solution of the Navier-Stokes equations for a porous channel with orthogonally moving walls. Phys. Fluids 15, 1485 (2003). https://doi.org/10.1063/1.1567719
El-Genk, M., Yang, I.: Numerical analysis of laminar flow in micro-tubes with a slip boundary. Energy Convers. Manag. 50(6), 1481–1490 (2009). https://doi.org/10.1016/j.enconman.2009.02.022
Farooq, J., Chung, J.D., Mushtaq, M., Lu, D., Ramazan, M., Farooq, U.: Influence of slip velocity on the flow of viscous fluid through a porous medium in a permeable tube with a variable bulk flow rate. Results Phys. 11, 861–868 (2018). https://doi.org/10.1016/j.rinp.2018.10.049
Guo, H., Gui, C., Lin, P., Zhao, M.: Multiple solutions and their asymptotics for laminar flows through a porous channel with different permeabilities. IMA J. Appl. Math. 85, 280–308 (2020). https://doi.org/10.1093/imamat/hxaa006
Kohler, J.P.: An investigation of laminar How through a porous-walled channel, Ph.D. Thesis, University of Massachusetts, Amherst (1973)
Kuo, U.U., Wang, C.A.: Multiple solutions of the steady flows in a rectangular channel with slip effect on two equally porous walls. Taiwan. J. Math. 16(3), 885–900 (2012). https://doi.org/10.11650/twjm/1500406663
Majdalania, J., Zhoua, C., Dawson, C.A.: Two-dimensional viscous flow between slowly expanding or contracting walls with weak permeability. J. Biomech. 35, 1399–1403 (2002). https://doi.org/10.1016/S0021-9290(02)00186-0
Michaels, A.S.: Ultrafiltration. In: Perry, E.S. (ed.) Progress in Separartion and Purification, vol. I. Interscience, New York (1968)
Nazari Moghaddam, R., Jamiolahmady, M.: Slip flow in porous media. Fuel 173, 298–310 (2016). https://doi.org/10.1016/j.fuel.2016.01.057
Nishiyama, N., Yokoyama, T.: Permeability of porous media: Role of the critical pore size. J. Geophys. Res. Solid Earth 122(9), 6955–6971 (2017)
Rasoulzadeh, M., Panfilov, M.: Asymptotic solution to the viscous/inertial flow in wavy channels with permeable walls. Phys. Fluids 30(10), 106604 (2018). https://doi.org/10.1063/1.5041748
Saffman, P.G.: On the boundary condition at the surface of a porous medium. Stud. Appl. Math. 50, 93–101 (1971). https://doi.org/10.1002/sapm197150293
Siddiqui, A.M., Gawo, G.A., Maqbool, K.: On slip of a viscous fluid through proximal renal tubule with linear reabsorption. AIMS Biophys. 7(1), 80–102 (2020). https://doi.org/10.3934/biophy.2021006
Singh, R., Laurence, R.L.: Influence of slip velocity at a membrane surface on ultrafiltration performance-I, channel flow system. Int. J. Heat Mass Transf. 22(5), 721–729 (1979). https://doi.org/10.1016/0017-9310(79)90119-4
Skjetne, E., Auriault, J.L.: Homogenization of wall-slip gas flow through porous media. Transp. Porous Media 36, 293–306 (1999). https://doi.org/10.1023/A:1006572324102
Sparrow, E., Beavers, G., Masha, B.: Laminar flow in a rectangular duct bounded by a porous wall. Phys. Fluids 17(7), 1465 (1974). https://doi.org/10.1063/1.1694915
Taherinejad, M., Afrouzan, A., Derakhshan, S.: CFD investigation of near-membrane slippery condition effects on water/salt transport in a reverse osmosis feed channel. Arab. J. Sci. Eng. 46(7), 6673–6685 (2021). https://doi.org/10.1007/s13369-020-05198-8
Ullah, H., Siddiqui, A.M., Sun, H., Haroon, T.: Slip effects on creeping flow of slightly non-Newtonian fluid in a uniformly porous slit. J. Braz. Soc. Mech. Sci. Eng. 41(10), 412 (2019). https://doi.org/10.1007/s40430-019-1917-2
Ullah, H., Lu, D., Siddiqui, A.M., Haroon, T., Maqbool, K.: Hydrodynamical study of creeping maxwell fluid flow through a porous slit with uniform reabsorption and wall slip. Mathematics 8(10), 1–22 (2020). https://doi.org/10.3390/math8101852
Ullah, H., Lu, D., Majeed Siddiqui, A., Maqbool, K., Iqbal, S.: Fluid model using recursive approach: application to permeable slit with uniform reabsorption and velocity slip. Results Phys. 25, 104196 (2021). https://doi.org/10.1016/j.rinp.2021.104196
Varunkumar, M., Muthu, P.: Fluid flow and solute transfer in a permeable tube with influence of slip velocity. Discontinuity Nonlinearity Complex. 9(1), 153–166 (2020). https://doi.org/10.5890/DNC.2020.03.011
Wang, C.A., Hwang, T.W., Chen, Y.Y.: Existence of solutions for Berman’s equation from laminar flows in a porous channel with suction. Comput. Math. Applic. 20(2), 35–40 (1990). https://doi.org/10.1016/0898-1221(90)90238-F
Zeidler, E.: Nonlinear functional analysis and its applications I: fixed-point theorems. Translated from the German by Wadsack, P.R. New York: Springer, ISBN: 978-0-387-90914-1 (1986)
Funding
The authors declare that no funds, grants, or other support were received during the preparation of this manuscript. Open Access funding enabled and organized by CAUL and its Member Institutions.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors have no relevant financial or non-financial interests to disclose.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Hao, M.L., Tisdell, C.C. When is the Porous, Laminar Flow Problem with Slip Condition Well Posed? And Where Does the Solution Lie?. Transp Porous Med 147, 281–303 (2023). https://doi.org/10.1007/s11242-023-01907-7
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11242-023-01907-7
Keywords
- Laminar flow
- Channel with porous walls
- Slip condition
- Solutions
- Contraction mapping
- Boundary value problem