Numerical solutions of the Falkner-Skan equation based on quasilinearization
|
|
- Leo Randall
- 5 years ago
- Views:
Transcription
1 Numerical solutions of the Falkner-Skan equation based on quasilinearization Shengfeng Zhu a,, Xiaoliang Cheng a, Qingbiao Wu a a Department of Mathematics, Zhejiang University, Hangzhou 327, Zhejiang, P.R.China Abstract We present two iterative methods for solving the Falkner-Skan equation based on the quasilinearization method. We formulate the original problem as a new free boundary value problem. The unknown free boundary depending on a small parameter is determined in iteratively. We apply the quasilinearization method to solve the nonlinear equation. Then we propose two different iterative approaches based on a cubic spline solver. The results of numerical experiments for various instances are compared with those reported previously in the literature. The comparisons show the accuracy, robustness and efficiency of the presented methodology. Key words: Falkner-Skan equation; Free boundary value problem; Quasilinearization method; Cubic spline; Semi-infinite interval Introduction Generally, no analytical solutions are available for the nonlinear two-point boundary value problems (BVPs). Numerical solutions of these problems have always been of interest for scientists and engineers. The well-known nonlinear third-order Falkner-Skan equation is much more challenging since it is a BVP depicted on an infinite interval. This problem arises in the study of laminar boundary layers exhibiting similarity in fluid mechanics []. The solutions of the Falkner-Skan equation are similarity solutions of the two-dimensional incompressible laminar boundary layer equations. The Falkner-Skan equation is given by f + βff + γ[ (f ) 2 ] =, < η < () Corresponding author. address: shengfengzhu@zju.edu.cn (Shengfeng Zhu). Preprint submitted to Applied Mathematics and Computation 2 August 29
2 subject to the boundary conditions f () = f () =, f ( ) =, (2) (3) where β and γ are constants and the prime stands for differentiation with respect to η. We can deduce easily that the solution of Eq. () satisfies the asymptotic condition f (η) as η. (4) Sometimes, the Falkner-Skan equation especially refers to the case when β =. The first analytical treatment for the Falkner-Skan equation was given by Hartree [2], who found that in the region.9884 < γ <, there exists a family of unique solutions whose first order derivative f (η) tends to exponentially. Weyl [3] proved that, for γ >, the problem has a solution f(η) whose first derivative f (η) increases with η and whose second derivative f (η) tends decreasingly to zero as η. Coppel [4] presented an elegant proof of the existence and uniqueness for γ > and showed that the wall shear stress α := f () is an increasing function of γ. In [5,6], extension for γ < was discussed and properties of solutions were investigated further. New theoretical results for the solutions of the Falkner-Skan equation were given by Padé [7] and Yang [8]. Numerical techniques for the solution of the Falkner-Skan equation fall into the following classes in the literature: finite difference [2,5,9 ], shooting [2 7], finite element [8], group invariance theory [9,2], Chebyshev spectral method [2,22], differential transformation method [23], etc. One problem for numerically solving the Falkner-Skan equation is how to deal with the infinite boundary. In the literature, there are various ways to overcome this difficulty. In [2,,4], a finite prescribed truncated boundary instead of infinity was introduced directly. Asaithambi [9,2,3,8] truncated the infinite boundary by an unknown η. The value for η is determined by imposing an asymptotic condition f (η ) =, which is an approximation of the condition (4). Fazio [2] proposed a novel approach to define the truncated boundary depending on a small parameter and formulate the Falkner- Skan equation as a free boundary value problem (FBVP). Recently, methods without truncating the infinite domain are developed, such as homotopy analysis method [24] and Adomian decomposition method [25], both of which aims to obtain series solutions and employ Padé approximation to enlarge the convergence domain and improve the accuracy of solutions. We apply the treatment for the infinite interval proposed by Fazio [9,2] and solve instead a FBVP as an approximation to the Falkner-Skan equation. Consider the FBVP resulting from the Falkner-Skan equation: 2
3 f + βf f + γ[ (f ) 2 ] =, f () =, f () =, f (η ) =, f (η ) =, (5) where <. The free boundary η can be considered as an unknown truncated boundary, which should be determined as part of solution. According to Theorem [2], the solution of FBVP (5) converges uniformly to the solution of Falkner-Skan equation on arbitrarily large interval [, η ] and the order of convergence is linear in at least. This result ensures theoretically the rationality for solving the FBVP instead of the original problem. The free truncated boundary η, which is dependent on an arbitrary small parameter, converges to as tends to. Therefore, this treatment is more reasonable than other direct truncation methods, such as [9,2 4,8,2]. We wish to compute approximately the solution of the FBVP, which is still nonlinear. Mandelzweig et al. [26] applied a powerful iteration technique called quasilinearization method (QLM) to nonlinear ordinary differential equations in physics. As a generalization of Newton-Raphson method, the QLM s iterations are constructed to yield quadratic and monotonic convergence. We employ the QLM to transform FBVP (5) to a sequence of quasilinear third-order differential equation BVPs, which can be solved much easier than the original FBVP. The remainder of the paper is organized as follows. In Section 2, using linear mappings, we transform the FBVP to a problem defined on [, ]. The QLM is employed to obtain the quasilinear form of the problem. Then we split it into two subproblems: a first-order initial value problem (IVP) and a second-order BVP. Section 3 is devoted to present a useful cubic spline solver for a class of two-point BVPs. In Section 4, we propose two different alternating iterative methods to solve the two problems. In every iteration, they both solve the second-order BVP by virtue of the cubic spline solver, but they differ in solving the IVP analytically or numerically. Then we present two corresponding algorithms. In Section 5, we apply our method to various cases of the Falkner-Skan equation by varying the values of β and γ. Finally, we use a brief conclusion to end the paper in Section 6. 2 Application of the QLM to the FBVP In this section, we will apply the QLM [26] to the problem (5) so as to linearize the FBVP. In order to change the physical domain [, η ] to a shorter interval [, ], we use the linear mappings x = η/η and g = f /η in Ref. [8] to the FBVP. Then, (5) is transformed to g + β η 2 gg + γη 2 [ (g ) 2 ] =, (6) g() =, g () =, g () =, (7) g () = η. (8) 3
4 Considering the nonlinearity of problem (6)-(8), we solve it iteratively both for g(x) and η. We are interested in obtaining a proper value η such that the solution g(x) of (6)-(7) would satisfy the boundary condition (8), i.e., u(η ) =, (9) where u(η ) := g (; η ) η and g (; η ) denotes the value g () is dependent on η. Concretely, we will realize our idea in a nested loop. Our nested iteration framework consists of an outer iteration for η and an inner iteration for g(x). We denote by η (k) the kth iterate of η. Let g (k,n) be the nth iterate for g (k) corresponding to η (k). For a given η (k), we can apply the QLM to the nonlinear problem (6)-(7) and obtain the n + th iterative approximation g (k,n+) (x) to the solution g (k) (x) in the inner iteration. More precisely, g (k,n+) (x) is obtained by solving the BVP g (k,n+) + βη (k)2 g (k,n) g (k,n+) 2γη (k)2 + βη (k)2 g (k,n) ( g (k,n+) g (k,n)) + γη (k)2 g (k,n+) () =, g (k,n+) () =, g (k,n+) () =, g (k,n) g (k,n+) [ (g (k,n) ) 2 + ] =, () where g (k,n) := d dx g(k,n), g (k,n) := d2 dx 2 g(k,n), g (k,n+) := d3 dx 3 g(k,n+), η (k)2 := (η (k) ) 2 and n =,, 2,.... According to [26], the zeroth iteration g (k,) (x) should satisfy at least one of the boundary conditions. A successful initial guess can generate quadratic convergence for the quasilinear iteration sequence {g (k,n) (x)} n=. See [26] for the detailed proof. In every iteration, such a third-order two-point BVP () can be solved analytically for g (k,n+) with Green s function. However, the solution procedure is rather complicated after a few iterations. Thus, here we try to evaluate g (k,n+) numerically instead. First, we split it into two subproblems. Set y (k,n) = g (k,n), () is transformed to a first-order IVP g (k,n) = y (k,n), g (k,n) () =, () and a second-order two-point BVP 4
5 y (k,n+) + βη (k)2 g (k,n) y (k,n+) 2γη (k)2 y (k,n) y (k,n+) + βη (k)2 y (k,n) (g (k,n+) g (k,n) ) [ + γη (k)2 (y (k,n) ) 2 + ] =, (2) y (k,n+) () =, y (k,n+) () =. In the inner iteration, if y (k,n) is known, () can be solved either analytically by direct integration or approximated by numerically methods for such first-order IVPs. Considering the smooth property of spline approximation, we shall use a cubic spline solver described in the following section to solve the linear problem (2). 3 A cubic spline solver This section is devoted to present a cubic spline solver for a class of two-point BVPs. Albasiny and Hoskins [27] derived the cubic spline approximation to a class of two-point BVPs y + p(x)y + q(x)y = r(x), (3) y() = a, y() = b. (4) Consider the uniform grid partition := { = x < x <... < x N = } of the interval [, ] with mesh size h = /N and the grid points x i = ih, i =,,..., N. The cubic spline S(x) interpolating to the function y(x) at the grid points is given by the piecewise expression S (x), if x [x, x ], S(x) = S i (x), if x [x i, x i ], i = 2, 3,..., N, S N (x), if x [x N, x N ], (5) where (x i x) 3 (x x i ) 3 S i (x) =M i + M i + 6h 6h ) ( + (y i h2 x 6 M xi i h (y i h2 6 M i ), x [x i, x i ], ) ( ) xi x h (6) for i =, 2,..., N; M i := S (ih); y i := y(x i ). Hence, S (x + i ) = h 3 M i h 6 M i+ + y i+ y i, i =,,..., N (7) h 5
6 and S (x i ) = h 3 M i + h 6 M i + y i y i, i =, 2,..., N. (8) h So the continuity of first derivatives at grid points implies that h 6 M i 2h 3 M i + h 6 M i+ = y i 2y i + y i+, i =, 2,..., N. (9) h Then, by Eq. (7), (8) and (9), the requirement that the spline approximation should satisfy the differential equation (3) at the grid points leads to a set of relationships from which we can eliminate M, M,..., M N. More precisely, using an approach similar to one in [27], we obtain a tridiagonal system of equations written as where E i y i F i y i + G i y i+ = H i i =, 2,..., N, (2) ( E i = d i h ) 2 p i + h2 6 q i, F i = c i ( + h ) ( 2 p i+ + d i h ) 2 p i 2h2 ( G i = c i + h ) 2 p i+ + h2 6 q i+, H i = h2 6 (d ir i + 4e i r i + c i r i+ ), c i = h 3 p i + h 3 p i h2 2 p i p i, d i = h 3 p i + h 3 p i+ h2 2 p ip i+, e i = h2 2 p i p i+ + 7h 24 (p i+ p i ), for i =, 2,..., N and 3 q ie i, p i = p(x i ), q i = q(x i ), r i = r(x i ) i =,,..., N. The boundary conditions (4) will correspond to y = a, y N = b. (2) 6
7 Thus, substituting (2) into the first and last equation, the tridiagonal system (2) can be solved fast and efficiently by the famous Thomas algorithm, which is a direct method. Denote and p := [p, p,..., p N ] T, q := [q, q,..., q N ] T, r := [r, r,..., r N ] T, y := [y, y,..., y N ] T, y := [y, y,..., y N] T, M := [M, M,..., M N ] T. When y is determined, we can calculate M by the formulae and M i = a i { ( + hp i 3 )[ r i q i y i p i h (y i y i ) ] + hp i 6 [ ri q i y i p i h (y i y i ) ]}, i =, 2,..., N (22) M i+ = b i { ( hp i 3 )[ r i+ q i+ y i+ p i+ h (y i+ y i ) ] hp i+ 6 [ ri q i y i p i h (y i+ y i ) ]}, i =,,..., N. Thus, the expression of cubic spline function S(x) on [, ] is available. It approximates y(x) at all points in [, ] to O(h 4 ) if y C 4 ([, ]). Since y i := y (x i ) = S (x i ) for i =,,..., N, we get y from (7) and (8). Remark M and M N should be evaluated by (22) and (23), respectively, while M, M 2,..., M N could be obtained either by (22) or (23). Similarly, y and y N should be evaluated by (7) and (8), respectively, while y, y 2,..., y N could be obtained either by (7) or (8). Remark 2 Based on the analysis above, we can build without difficulty a cubic spline solver in the form of a function CubicSplineBVP (p, q, r, a, b) for solving the twopoint BVP class (3)-(4). The output are the vectors y, y and M, from which we also obtain the expression of S(x). (23) 4 Two iterative methods based on a cubic spline solver In this section, we will propose two different alternating iterative methods to solve the problem consisting of () and (2) by virtue of the cubic spline solver depicted above. 7
8 Let y (k,n) i = y (k,n) (x i ) for i =,,..., N. Denote y (k,n) := [ y (k,n), y (k,n),..., y (k,n) ] T, N y (k,n) := [ y (k,n), y (k,n)..., y (k,n) ] T, N and g (k,n) := [ g (k,n), g (k,n),..., g (k,n) ] T. N All the notations here and in the following stand for the same quantities as those appearing before except for the two superscripts. We now describe the alternating iterative methods for solving these two subproblems. For (), if y (k,n) is an integrable function defined on [, ], analytical treatment leads to g (k,n) = x y(k,n) (t) dt from (). Hence, g (k,n) xi i = y (k,n) (x) dx (24) for i =, 2,..., N. Thus, g (k,n) is available if y (k,n) is known. Alternatively, for (), we may adopt a numerical integration method, such as Heun s method. Then we have g (k,n) =, g (k,n) i+ = g (k,n) i + h 2 ( y (k,n) i + y (k,n) ) (25) i+, for i =,,..., N. Therefore, g (k,n) is available if y (k,n) is known. Remark 3 Heun s method used here is an explicit scheme. It is equivalent to the central difference discretization of () at x = x i+/2 for i =,,..., N (see [9,8]). For (2), when g (k,n) is already known by (24) or (25), it is still difficult to construct iteration schemes from y (k,n) to y (k,n+) in Eq. (2) directly since y (k,n) and g (k,n+) are unknown. We use an intermediate iteration step to obtain a function g (k,n+ 2 ) as an approximation to g (k,n+). To realize it, we first introduce three problems denoted by Q, Q 2 and Q 3 as follows. Q y (k,n+ 2 ) + βη (k)2 g n y (k,n+ 2 ) 2γη (k)2 y (k,n+ 2 ) () =, y (k,n+ 2 ) () =, [ y (k,n) y (k,n+ 2 ) + γη (k)2 (y (k,n) ) 2 + ] =, (26) Q 2 g (k,n+ 2 ) = y (k,n+ 2 ), g (k,n+ 2 ) () =, (27) 8
9 Q 3 y (k,n+) + βη (k)2 + βη (k)2 g (k,n) y (k,n+) 2γη (k)2 y (k,n+) () =, y (k,n+) () =. y (k,n) ( g (k,n+ 2 ) g (k,n)) + γη (k)2 y (k,n) y (k,n+) [ (y (k,n) ) 2 + ] =, (28) In the following, we approximate Q, Q 2 and Q 3 successively. Note that the forms of Q and Q 3 belong to the BVP class (3)-(4). Thus, Q and Q 3 can be solved by the cubic spline solver CubicSplineBVP. For Q, p k,n (x) = βη (k)2 g (k,n), q k,n (x) = 2γη (k)2 [ y (k,n), r k,n (x) = γη (k)2 (y (k,n) ) 2 + ]. For Q 3, p k,n (x) = p k,n (x), q k,n (x) = q k,n (x), r k,n (x) = r k,n (x) βη (k)2 y (k,n) ( g (k,n+ 2 ) g (k,n)). Now we present the iteration procedure from g (k,n) to y (k,n+). First, the cubic spline approximation S (k,n+ 2 ) (x) for Q, y (k,n+ 2 ), y (k,n+ 2 ) and M (k,n+ 2 ) are all obtained by calling the function CubicSplineBVP (p k,n, q k,n, r k,n,, ). Second, g (k,n+ 2 ) can be obtained analytically by (24) or numerically by (25). If we want to solve Q 2 analytically, we have g (k,n+ 2 ) xi i = y (k,n+ 2 ) (x) dx i xj S (k,n+ 2 ) j j= x j (x) dx. Moreover, using (5), (22) and (23), we obtain g (k,n+ 2 ) i = i j= xj x j [ M (k,n+ 2 ) (x j x) 3 j 6h h2 6 M (k,n+ 2 ) )( x j x) ( (k,n+ j + y 2 ) j h i [ h 3 = j= 24 M (k,n+ 2 ) j + h3 24 M (k,n+ 2 ) j + h 2 + h ( (k,n+ y 2 ) j h2 2 6 M (k,n+ 2 ) ) ] j + M (k,n+ 2 ) (x x j ) 3 j + ( y (k,n+ 2 ) j 6h 6 M (k,n+ 2 ) )( x x ) ] j j dx h h2 6 M (k,n+ 2 ) ) j h2 ( y (k,n+ 2 ) j (29) 9
10 = h3 24 i j= ( (k,n+ M 2 ) j + M (k,n+ 2 ) ) h j + 2 i j= ( (k,n+ y 2 ) j + y (k,n+ 2 ) ) j for i =, 2,..., N. If we approximate Q 2 by Heun s method, we get g (k,n+ 2 ) =, g (k,n+ 2 ) i+ = g (k,n+ 2 ) i + h 2 for i =,,..., N. ( (k,n+ y 2 ) i + y (k,n+ 2 ) ) (3) i+, Last, if g (k,n+ 2 ) is available by (29) or (3), the cubic spline approximation S (k,n+) (x) for Q 2, y (k,n+), y (k,n+) and M (k,n+) are all obtained by calling the function Cubic- SplineBVP ( p k,n, q k,n, r k,n,, ). In the next iterative step, g (k,n+) can be obtained, similar as (29) or (3), by g (k,n+) i = h3 24 i j= ( M (k,n+) j + M (k,n+) ) h j + 2 i j= ( y (k,n+) j + y (k,n+) ) j (3) for i =, 2,..., N or g (k,n+) =, g (k,n+) i+ = g (k,n+) i + h 2 for i =,,..., N. ( y (k,n+) i + y (k,n+) ) (32) i+ For solving Q, Q 2 and Q 3 iteratively, the iterations described above will continue until y (k,n+) y (k,n) < ε or g (k,n+) g (k,n) < ε (33) for some prescribed error tolerance ε. Now we discuss the outer iteration loop for η. We assume that condition (33) is satisfied corresponding to some k and denote g (k) and ỹ (k) the converged values for g (k,n) and y (k,n), respectively. η may be calculated by solving the nonlinear Eq. () by means of an iterative technique such as Newton s method or secant method. Since u(η ) and its derivative are not known explicitly, we will use secant method for Eq. () to obtain iterate for η. First we have u(η (k) ) = g (; η (k) ) η (k) = y (; η (k) ) η (k) (34)
11 where y (; η (k) ) = ỹ (k) N = h 3 Then, the secant iterate is obtained using M N k + h M N k + ỹk N ỹn k. (35) 6 h η (k+) = η (k) + η (k), (36) where η (k) = u(η (k) ) η (k) u ( η (k) The iterative procedure is terminated when the condition η (k ) ) ( ). (37) (k ) u η u(η (k) ) < ε2 (38) is satisfied for some prescribed tolerance ε 2. We will assume that we begin our computations with initial guesses η () and η () for η. Now we are ready to present the algorithms of our numerical methods for the Falkner- Skan equation. In our iterative procedure, solving first-order IVPs analytically or numerically leads to two different algorithms below. Algorithm Step. Input, ε, ε 2, η (), η (), h, and an initial function y (,) (x); N /h. Step 2. k, repeat through step 8 until (38) is satisfied. Step 3. n, {y (k,n) i } N i= {y (,) (x i )} N i=, fix y and y N ; repeat through step 7 until (33) is satisfied. Step 4. If n =, compute g (k,n) using (24); else compute g (k,n) using (3). Step 5. Obtain y (k,n+ 2 ) and M (k,n+ 2 ) by CubicSplineBVP (p k,n, q k,n, r k,n,, ). Step 6. Compute g (k,n+ 2 ) by (29). Step 7. Obtain y (k,n+), y (k,n+) and M (k,n+) by CubicSplineBVP ( p k,n, q k,n, r k,n,, ); n n +. Step 8. Compute u(η (k) ), η (k) and η (k+) by (34) and (36); k k +. Algorithm 2 Step. Input, ε, ε 2, η (), η (), h and an initial vector y (,) with y (,) ; N /h. y (,) N Step 2. k, repeat through step 8 until (38) is satisfied. and
12 Step 3. n, y (k,n) y (,) ; repeat through step 7 until (33) is satisfied. Step 4. Compute g (k,n) using (25); Step 5. Obtain y (k,n+ 2 ) by CubicSplineBVP (p k,n, q k,n, r k,n,, ). Step 6. Compute g (k,n+ 2 ) by (3). Step 7. Obtain y (k,n+) and y (k,n+) by CubicSplineBVP ( p k,n, q k,n, r k,n,, ); n n +. Step 8. Compute u(η (k) ), η (k) and η (k+) by (34) and (36); k k +. Remark 4 In Algorithm 2, Heun s method could be replaced by other high order methods such as Runge-Kutta method to improve accuracy. Hereafter, we use the tilde notation placed on top of a symbol to indicate its corresponding final converged quantity. Let S(x) be the final converged cubic spline approximation obtained by Algorithm or Algorithm 2. Then we get its expression as follows: g i (x) = g i + x x i Si (t) dt = g i + h2 (x i x) 2 2h [h2 (x i x) 2 ] 2 24h = g i + ỹi 2 h M i 6ỹ i M i h 2 (x x i ) 2 + 2h ỹ i + (x x i ) 2 ỹ i 2h M i + (x x i ) 4 2h 2 (x x i ) 2 24h 24 h3 + 6ỹ i M i h 2 (x x i ) 2 2h M i 24h (x x i ) 4 M i M i 24h (x x i) 4 (39) x [x i, x i ], for i =, 2,..., N. Let g (x), if x [x, x ], g(x) = g i (x), if x [x i, x i ], i = 2, 3,..., N, g N (x), if x [x N, x N ], (4) then we have f(η) f (η) = η g(x) = η g(η/ η ). (4) Notice that the final expression (4) is a fourth degree polynomial approximation for the Falkner-Skan equation. Consequently, by (4), we get 2
13 f (η) f (η) = g (η/ η ), (42) η f (η i ) g i (x i ) η = η ( h 3 M i h 6 M i + ỹi ỹ i h for i =, 2,..., N and obtain the wall shear stress α approximately by where α := f () = η ( h 3 ), (43) M h ) M + ỹ ỹ, (44) 6 h M i = {( a i+ + hp i 6 ) [ r i q i ỹ i p ] i h (ỹ i+ ỹ i ) + hp i+ 3 [ r i+ q i+ ỹ i+ p i+ h (ỹ i+ ỹ i ) ] }, i =,. (45) Remark 5 Since g () = S (), by (5), we can also obtain the formula for computing α exactly the same as (44). 5 Numerical results In this section, numerous cases of the Falkner-Skan equation are solved numerically by our methods. We observe, through many experiments, that Algorithm 2 has almost the same accuracy as Algorithm but is more efficient than Algorithm in implementation. This is also why we use Heun s method rather than Runge-Kutta method in Algorithm 2. For the sake of simplicity, we only list numerical results obtained by Algorithm 2. Comparisons are made between our numerical results and those reported previously in the literature. 5. Case : Blasius flow(β = /2, γ = ) When β = /2 (or β = ) andγ =, we have the so-called Blasius flow. Blasius [28] obtained this problem in the study of a laminar boundary layer along a thin flat plate. The Blasius flow has been either studied individually in [29 3] and the references therein or tested as a special case of Falkner-Skan equation in [3,32], etc. Generally, in order to obtain an idea of the accuracy involved, the value of the wall shear stress is of physical interest since it defines the skin friction around the plate. Table shows the comparison of the wall shear stress and the truncated boundary obtained by the proposed method to those obtained by [32]. We choose h = 3, ε = and = 3
14 ε 2, ε 2 for various ε 2. Denote α,h and η,h the values of wall shear stress and truncated boundary corresponding to and h, respectively. Then increased accuracy is accomplished by using extrapolation formulas α = (4α,h/2 α,h )/3 and η = (4η,h/2 η,h )/3. Note that the present results for α are also in good agreement with those reported in [3] and [29]. We estimate the order of accuracy for calculation of α by grid refinement analysis. Let α denote the exact value of α. Notice that the α for Blasius flow is unknown, we estimate the order of accuracy by ( ) α,h/2 α,h order = log 2. (46) α,h/4 α,h/2 Shown in Table 2 are results with ε = and ε 2 = 9. In Fig., we plot the curves of df / dη and d 2 f / dη 2 in cases = 7 and = 9, respectively. 5.2 Case 2: Pohlhausen flow(β =, γ = ) The exact solution of Pohlhausen [33] flow is known as f (η) = 3 tanh 2 ( η 2 + tanh 2 3 ) 2. Hence, α = 2/ 3 and the exact expression of f (η) is also available. We denote by E,h := α,h α the error in the computed solution corresponding to h and, then estimate the order of accuracy using the formula order = log(e,h/e,h/2 ). (47) log 2 Summarized in Table 3 are results obtained with ε = and ε 2 = 9. We clearly see the second-order accuracy of our method. In Fig. 2, comparisons of f (η) and f (η) are made between exact values with our numerical results. 5.3 Case 3: Several instances(β = ) The Falkner-Skan equation mostly refers to the case corresponding to β =. The solutions corresponding to γ >, γ =, and γ < are known as accelerating, constant, and decelerating flows, respectively. 4
15 5.3. Homann flow(β =, γ = /2) The numerical results obtained for the Homman flow, corresponding to the value of β =, γ = /2, are listed in Table 4. We compare our results with those reported in Fazio [2] using ε = and ε 2 = 8. In Fig. 3, we plot the curves of df / dη and d 2 f / dη 2 when = 6 and = Hiemenz flow(β =, γ = ) The Falkner-Skan equation turns to the Hiemenz flow when β = and γ = [34]. In Fig. 4, we plot the curves of df / dη and d 2 f / dη 2 when = 6 and = 7. Finally, the scaling curves f /η of f for the above four special flows are presented in Fig Other cases (β =, γ [.988, ]) To illustrate the robustness and accuracy of our algorithms further, we perform many other tests for the cases that β = and γ [.988, ]. The present results in Table 5, Fig. 6 and Fig. 7 were obtained with = 6, ε =, and ε 2 = 7. Our calculated α are in good agreement with those reported [3] and [32]. We also observe that the free truncated boundary η is more accurate than [32] especially as 6. Fig. 8 shows the profile of α as an increasing function of γ, namely, we verify numerically Coppel [4] s theoretical result. The figure was obtained with ε =, and ε 2 = Computational effort The amount of computational effort needed mainly depends on ε and ε 2. All the computations are carried out on a PC with 3. GHz Pentium 4 CPU and G memory. each instance usually costs 5-3 ms. All numerical results are obtained with η () = and η () = 2 for all cases. In Table 6, we list the number of quasilinearization iterations in inner loop for different flows of Blasius, Pohlhausen, Homman and Hiemenz as test examples. Denote by N and N 2 the minimal and maximal number of quasilinearization iterations, respectively. Our results are obtained when = 7, ε 2 = 8 and {y (,) j } N j= =, y (,) N =. We find that N and N 2 are both independent of and ε 2 through various experiments. In Table 7, we report the number of outer iterations for η. We choose flows of Blasius, Pohlhausen, Homman and Hiemenz as test examples. We fix ε = as we find the iteration number is independent of ε 2. Denote by K, K 2 and K 3 the number of outer iterations corresponding to ε 2 = 8, ε 2 = 9 and ε 2 =, respectively. 5
16 6 Conclusions We have presented two iterative methods for solving the Falkner-Skan equation based on the quasilinearization technique and a cubic spline solver. We formulate the original problem as a new FBVP. The truncated unknown free boundary depending on a small parameter is determined in iteratively. The quasilinearization method gives a sequence of linear differential equations. Every linear third-order BVP is split into a IVP and a second-order BVP. Solving the IVP analytically or numerically leads to two different iterative methods. The resulting two algorithms both have a nested iteration framework. In every inner iterative step, the cubic spline solver is employed to solve an intermediate step second-order BVP. The numerical results obtained with the second algorithm are presented for various instances of the Falkner-Skan equation. The comparisons between our results with those obtained by others illustrate the accuracy, robustness and efficiency of our methodology. Acknowledgements This work was supported by National Natural Science Foundation of China (No. 8778) and Natural Science Foundation (No. Y6654) of Zhejiang Province. References [] V.M. Falkner, S.W. Skan, Some approximations of the boundary layers equations, Philos. Mag. Soc. 2 (93) [2] D.R. Hartree, On an equation occurring in Falkner-Skan approximate treatment of the equations of the boundary layer, Proc. Camb. Phil. Soc. 33 (937) [3] H. Weyl, On the differential equations of the simplest boundary-layer problem, Ann. Math. 43 (942) [4] W.A. Coppel, On a differential equation of boundary layer theory, Philos. Trans. Roy. Soc. London Ser. A 253 (96) -36. [5] B. Oskam, A.E.P. Veldman, Branching of the Falkner-Skan solutions for λ <, J. Engg. Math. 6 (982) [6] A.E.P. Veldman, A.I. Van de Vooren, On a generalized Falkner-Skan equation, J. Math. Anal. Appl. 25 (98) 2-. [7] O. Padé, On the solution of Falkner-Skan equation, J. Math. Anal. Appl. 285 (23) [8] G.C. Yang, New results of Falkner-Skan equation arising in boundary layer theory, Appl. Math. Comput. 22 (28)
17 [9] A. Asaithambi, A finite-difference method for the solution of the Falkner-Skan equation, Appl. Math. Comput. 92 (998) [] A. Asaithambi, A second-order finite-difference method for the Falkner-Skan equation, Appl. Math. Comput. 56 (24) [] P.M. Beckett, Finite difference solution of boundary-layer type equations, Int. J. Comput. Math. 4 (983) [2] A. Asaithambi, A numerical method for the solution of the Falkner-Skan equation, Appl. Math. Comput. 8 (997) [3] A. Asaithambi, Solution of the Falkner-Skan equation by recursive evaluation of Taylor coefficients, J. of Comput. Appl. Math. 76 (25) [4] T. Cebeci, H.B. Keller, Shooting and parallel shooting methods for solving the Falkner-Skan boundary-layer equation, J. Comp. Phys. 7 (97) [5] C.W. Chang, J.R. Chang, C.S. Liu, The Lie-group shooting method for boundary layer equations in fluid mechanics, J. Hydrodynamics Ser.B 8 (26) 3-8. [6] C.S. Liu, J.R. Chang, The Lie-group shooting method for multiple-solutions of Falkner-Skan equation under suctioninjection conditions, Int. J. Non-Linear Mech. 43 (28) [7] I. Sher, A.Yakhot, New approach to the solution of the Falkner-Skan equation, AIAA J. 39 (5) (2) [8] A. Asaithambi, Numerical solution of the Falkner-Skan equation using piecewise linear functions, Appl. Math. Comput. 59 (24) [9] R. Fazio, The Falkner-Skan equation: Numerical solutions within group invariance theory, Calcolo, 3 (994) [2] R. Fazio, A novel approach to the numerical solution of boundary value problems on infinite intervals, SIAM J. Numer. Anal. 33 (996) [2] E.M.E. Elbarbary, Chebyshev finite difference method for the solution of boundary-layer equations, Appl. Math. Comp. 6 (25) [22] H. Nasr, I.A. Hassanien, H.M. El-Hawary, Chebyshev solution of laminar boundary layer flow, Int. J. Comput. Math. 33 (99) [23] B.L. Kuo, Application of the differential transformation method to the solutions of the Falkner-Skan wedge flow, Acta Mech. 64 (23) [24] S.J. Liao, A non-iterative numerical approach for two-dimensional viscous flow problems governed by the Falkner-Skan equation, Int. J. Numer. Meth. Fluids 35 (2) [25] N.S. Elgazery, Numerical solution for the Falkner-Skan equation, Chaos Solit. Fract. 35 (28) [26] V.B. Mandelzweig, F. Tabakin, Quasilinearization approach to nonlinear problems in physics with application to nonlinear ODEs, Comput. Phys. Comm. 4 (2)
18 [27] E.L Albasiny, W.D. Hoskins, Cubic splines solutions to two-point boundary value problems, Comput. J. 2 (969) [28] H. Blasius, Grenzschichten in Flüssigkeiten mit kleiner Reibung, Z. Math. Phys. 56 (98) -37. [29] R. Fazio, The Blasius problem formulated as a free boundary value problem, Acta Mech. 95 (992) -7. [3] S.J. Liao, A uniformly valid analytic solution of two-dimensional viscous flow over a semiinfinite flat plate, J. Fluid Mech. 385 (999) -28. [3] L. Wang, A new algorithm for solving classical Blasius equation, Appl. Math. Comp. 57 (24) -9. [32] A.A. Salama, Higher order method for solving free boundary problems, Numer. Heat Transfer, Part B: Fundamentals 45 (24) [33] K. Pohlhausen, Zur näherungsweisen Integration der Differentialgleichung der laminaren Grenzschicht, J. Appl. Math. Mech. (ZAMM) (92) [34] K. Hiemenz, Die Grenzschicht an einem in den gleichförmigen Flüssigkeitsstrom eingetauchten geraden Kreiszylinder, Dingl. Polytech. J 326 (9)
19 Table Comparisons of truncated boundary η and the wall shear stress α for the Blasius equation. ε 2 = ε 2 = ε 2 α [32] η [32] α η α η Table 2 Estimating order of accuracy by grid refinement analysis when α is unknown. h α,h α,h/2 α,h/4 order / / / Table 3 Estimating order of accuracy by grid refinement analysis when α is known. h α,h E h α,h/2 E h/2 order / / /
20 Table 4 Comparison of the truncated boundary x ε and wall shear stress α for the Homann flow. Present α α [2] Present η η [2] Table 5 Comparison of the free boundary x ε and wall shear stress α when β =. γ Present α Asaithambi [3] α [32] Present η η [32] Table 6 The number of quasilinearization iterations. ε Blasius P ohlhausen Homman Hiemenz N N 2 N N 2 N N 2 N N
21 Table 7 The number of outer iterations for η. ε Blasius P ohlhausen Homman Hiemenz K K 2 K 3 K K 2 K 3 K K 2 K 3 K K 2 K df ε /dη (ε= 9 ).8 d 2 f ε /dη 2 (ε= 9 ) df ε /dη (ε= 7 ).6 d 2 f ε /dη 2 (ε= 7 ) η Fig.. Numerical solution of the Blasius flow. 2
22 Exact df/dη Exact d 2 f/dη 2 Numerical df ε /dη Numerical d 2 f ε /dη η Fig. 2. Comparison between numerical solution and exact solution for the Pohlhausen flow df /dη (ε= 7 ) ε d 2 f /dη 2 (ε= 7 ) ε df /dη (ε= 6 ) ε d 2 f ε /dη 2 (ε= 6 ) η Fig. 3. Numerical solution of the Homman flow. 22
23 df /dη (ε= 7 ) ε d 2 f /dη 2 (ε= 7 ) ε df ε /dη (ε= 6 ) d 2 f ε /dη 2 (ε= 6 ) η Fig. 4. Numerical solution of the Hiemenz flow Hiemenz Homman Blasius Pohlhausen.6 f ε /η ε η/η ε Fig. 5. Numerical f for 4 special flows. 23
24 γ=4 γ=3 γ=2 γ= γ= γ=8 γ=6 γ=4 df ε /dη.6.5 df ε /dη η (a) η (b) γ=2. γ=. γ=.5 γ= γ=. γ=.5 γ=.8 γ=.988 df ε /dη.6.5 df ε /dη η (c) η (d) Fig. 6. Numerical approximation of df/ dη corresponding to different values for γ. 24
25 ε γ=4 γ=3 γ=2 γ= γ= γ=8 γ=6 γ= d 2 f/dη 2 4 d 2 f/dη η η (a) (b) γ=2. γ=. γ=.5 γ= γ=. γ=.5 γ=.8 γ= d 2 f/dη d 2 f/dη η η (c) (d) Fig. 7. Numerical approximation of d 2 f/ dη 2 corresponding to different values for γ ε= 7 ε= γ Fig. 8. Profile of α as a function of γ. 25
FURTHER SOLUTIONS OF THE FALKNER-SKAN EQUATION
FURTHER SOLUTIONS OF THE FALKNER-SKAN EQUATION LAZHAR BOUGOFFA a, RUBAYYI T. ALQAHTANI b Department of Mathematics and Statistics, College of Science, Al-Imam Mohammad Ibn Saud Islamic University (IMSIU),
More informationTHE LIE-GROUP SHOOTING METHOD FOR BOUNDARY LAYER EQUATIONS IN FLUID MECHANICS *
Conference of Global Chinese Scholars on Hydrodynamics HE LIE-GROUP SHOOING MEHOD FOR BOUNDARY LAYER EQUAIONS IN FLUID MECHANICS * CHANG Chih-Wen CHANG Jiang-Ren LIU Chein-Shan Department of Systems Engineering
More informationTHE DIFFERENTIAL TRANSFORMATION METHOD AND PADE APPROXIMANT FOR A FORM OF BLASIUS EQUATION. Haldun Alpaslan Peker, Onur Karaoğlu and Galip Oturanç
Mathematical and Computational Applications, Vol. 16, No., pp. 507-513, 011. Association for Scientific Research THE DIFFERENTIAL TRANSFORMATION METHOD AND PADE APPROXIMANT FOR A FORM OF BLASIUS EQUATION
More informationPrandtl-Blasius Flow Steven Finch. November 12, 2008
The boundary value problem Prandtl-Blasius Flow Steven Finch November 12, 2008 000 ( )+ 00 ( ) ( ) =0 (0) = 0 0 (0) = 0 lim 0 ( ) =1 arises in the study of two-dimensional incompressible viscous flow past
More informationBoundary value problem with integral condition for a Blasius type equation
ISSN 1392-5113 Nonlinear Analysis: Modelling and Control, 216, Vol. 21, No. 1, 114 12 http://dx.doi.org/1.15388/na.216.1.8 Boundary value problem with integral condition for a Blasius type equation Sergey
More informationA Posteriori Error Estimator for a Non-Standard Finite Difference Scheme Applied to BVPs on Infinite Intervals
arxiv:1503.00037v2 [math.na] 19 Mar 2015 A Posteriori Error Estimator for a Non-Standard Finite Difference Scheme Applied to BVPs on Infinite Intervals Riccardo Fazio and Alessandra Jannelli Department
More informationFALLING FILM FLOW ALONG VERTICAL PLATE WITH TEMPERATURE DEPENDENT PROPERTIES
Proceedings of the International Conference on Mechanical Engineering 2 (ICME2) 8-2 December 2, Dhaka, Bangladesh ICME-TH-6 FALLING FILM FLOW ALONG VERTICAL PLATE WITH TEMPERATURE DEPENDENT PROPERTIES
More informationA NOTE ON BLASIUS TYPE BOUNDARY VALUE PROBLEMS. Grzegorz Andrzejczak, Magdalena Nockowska-Rosiak, and Bogdan Przeradzki
Opuscula Math. 33, no. 1 213, 5 17 http://dx.doi.org/1.7494/opmath.213.33.1.5 Opuscula Mathematica A NOTE ON BLASIUS TYPE BOUNDARY VALUE PROBLEMS Grzegorz Andrzejczak, Magdalena Nockowska-Rosiak, and Bogdan
More informationA curvature-unified equation for a non-newtonian power-law fluid flow
Int. J. Adv. Appl. Math. and Mech. 2(3) (215) 72-77 (ISSN: 2347-2529) Journal homepage: www.ijaamm.com International Journal of Advances in Applied Mathematics and Mechanics A curvature-unified equation
More informationCHAPTER-III GENERAL GROUP THEORETIC TRANSFORMATIONS FROM BOUNDARY VALUE TO INITIAL VALUE PROBLEMS
CHAPTER-III GENERAL GROUP THEORETIC TRANSFORMATIONS FROM BOUNDARY VALUE TO INITIAL VALUE PROBLEMS 3.1 Introduction: The present chapter treats one of the most important applications of the concept of continuous
More informationNon-unique solution for combined-convection assisting flow over vertical flat plate
Sādhanā Vol. 31, Part 6, December 2006, pp. 709 719. Printed in India Non-unique solution for combined-convection assisting flow over vertical flat plate K VENKATASUBBAIAH, AMRITA MITTAL and T K SENGUPTA
More information(f(x) P 3 (x)) dx. (a) The Lagrange formula for the error is given by
1. QUESTION (a) Given a nth degree Taylor polynomial P n (x) of a function f(x), expanded about x = x 0, write down the Lagrange formula for the truncation error, carefully defining all its elements. How
More informationTHE full Navier-Stokes equations are difficult or impossible
Unsteady Reversed Stagnation-Point Flow over a Flat Plate Vai Kuong Sin, Member, ASME; Fellow, MIEME, and Chon Kit Chio arxiv:130.997v1 [physics.flu-dyn] 13 Feb 013 Abstract This paper investigates the
More informationMHD stagnation-point flow and heat transfer towards stretching sheet with induced magnetic field
Appl. Math. Mech. -Engl. Ed., 32(4), 409 418 (2011) DOI 10.1007/s10483-011-1426-6 c Shanghai University and Springer-Verlag Berlin Heidelberg 2011 Applied Mathematics and Mechanics (English Edition) MHD
More informationComparison of homotopy analysis method and homotopy perturbation method through an evolution equation
Comparison of homotopy analysis method and homotopy perturbation method through an evolution equation Songxin Liang, David J. Jeffrey Department of Applied Mathematics, University of Western Ontario, London,
More informationUnsteady Stagnation-point Flow of a Second-grade Fluid
Avestia Publishing Journal of Fluid Flow, Heat Mass Transfer (JFFHMT) Volume 3, Year 016 ISSN: 368-6111 DOI: 10.11159/jffhmt.016.003 Unsteady Stagnation-point Flow of a Second-grade Fluid Fotini Labropulu
More informationNumerical solution of general boundary layer problems by the method of dierential quadrature
Scientia Iranica B (2013) 20(4), 1278{1301 Sharif University of Technology Scientia Iranica Transactions B: Mechanical Engineering www.scientiairanica.com Research Note Numerical solution of general boundary
More informationScaling and Spiral Equivalence from a Numerical Viewpoint
Scaling and Spiral Equivalence from a Numerical Viewpoint Riccardo Fazio and Salvatore Iacono Abstract The non-iterative numerical solution of nonlinear boundary value problems is a subject of great interest.
More informationChapter 2 Interpolation
Chapter 2 Interpolation Experiments usually produce a discrete set of data points (x i, f i ) which represent the value of a function f (x) for a finite set of arguments {x 0...x n }. If additional data
More informationExact Solutions for a Class of Singular Two-Point Boundary Value Problems Using Adomian Decomposition Method
Applied Mathematical Sciences, Vol 6, 212, no 122, 697-618 Exact Solutions for a Class of Singular Two-Point Boundary Value Problems Using Adomian Decomposition Method Abdelhalim Ebaid 1 and Mona D Aljoufi
More informationChapter 11 ORDINARY DIFFERENTIAL EQUATIONS
Chapter 11 ORDINARY DIFFERENTIAL EQUATIONS The general form of a first order differential equations is = f(x, y) with initial condition y(a) = y a We seek the solution y = y(x) for x > a This is shown
More informationSolution of a Fourth Order Singularly Perturbed Boundary Value Problem Using Quintic Spline
International Mathematical Forum, Vol. 7, 202, no. 44, 279-290 Solution of a Fourth Order Singularly Perturbed Boundary Value Problem Using Quintic Spline Ghazala Akram and Nadia Amin Department of Mathematics
More information(0, 0), (1, ), (2, ), (3, ), (4, ), (5, ), (6, ).
1 Interpolation: The method of constructing new data points within the range of a finite set of known data points That is if (x i, y i ), i = 1, N are known, with y i the dependent variable and x i [x
More informationResearch Article Analytic Solution for MHD Falkner-Skan Flow over a Porous Surface
Applied Mathematics Volume 01, Article ID 13185, 9 pages doi:10.1155/01/13185 Research Article Analytic Solution for MHD Falkner-Skan Flow over a Porous Surface Fatheah A. Hendi 1 and Majid Hussain 1 Department
More informationImproving convergence of incremental harmonic balance method using homotopy analysis method
Acta Mech Sin (2009) 25:707 712 DOI 10.1007/s10409-009-0256-4 RESEARCH PAPER Improving convergence of incremental harmonic balance method using homotopy analysis method Yanmao Chen Jike Liu Received: 10
More informationIntegration, differentiation, and root finding. Phys 420/580 Lecture 7
Integration, differentiation, and root finding Phys 420/580 Lecture 7 Numerical integration Compute an approximation to the definite integral I = b Find area under the curve in the interval Trapezoid Rule:
More informationThe family of Runge Kutta methods with two intermediate evaluations is defined by
AM 205: lecture 13 Last time: Numerical solution of ordinary differential equations Today: Additional ODE methods, boundary value problems Thursday s lecture will be given by Thomas Fai Assignment 3 will
More informationChapter 8. Numerical Solution of Ordinary Differential Equations. Module No. 1. Runge-Kutta Methods
Numerical Analysis by Dr. Anita Pal Assistant Professor Department of Mathematics National Institute of Technology Durgapur Durgapur-71309 email: anita.buie@gmail.com 1 . Chapter 8 Numerical Solution of
More informationApplied Math for Engineers
Applied Math for Engineers Ming Zhong Lecture 15 March 28, 2018 Ming Zhong (JHU) AMS Spring 2018 1 / 28 Recap Table of Contents 1 Recap 2 Numerical ODEs: Single Step Methods 3 Multistep Methods 4 Method
More informationAn Elegant Perturbation Iteration Algorithm for the Lane-Emden Equation
Volume 32 - No.6, December 205 An Elegant Perturbation Iteration Algorithm for the Lane-Emden Equation M. Khalid Department of Mathematical Sciences Federal Urdu University of Arts, Sciences & Techonology
More informationNUMERICAL SOLUTION OF MHD FLOW OVER A MOVING VERTICAL POROUS PLATE WITH HEAT AND MASS TRANSFER
Int. J. Chem. Sci.: 1(4), 14, 1487-1499 ISSN 97-768X www.sadgurupublications.com NUMERICAL SOLUTION OF MHD FLOW OVER A MOVING VERTICAL POROUS PLATE WITH HEAT AND MASS TRANSFER R. LAKSHMI a, K. JAYARAMI
More informationAn Overly Simplified and Brief Review of Differential Equation Solution Methods. 1. Some Common Exact Solution Methods for Differential Equations
An Overly Simplified and Brief Review of Differential Equation Solution Methods We will be dealing with initial or boundary value problems. A typical initial value problem has the form y y 0 y(0) 1 A typical
More informationChapter 5 HIGH ACCURACY CUBIC SPLINE APPROXIMATION FOR TWO DIMENSIONAL QUASI-LINEAR ELLIPTIC BOUNDARY VALUE PROBLEMS
Chapter 5 HIGH ACCURACY CUBIC SPLINE APPROXIMATION FOR TWO DIMENSIONAL QUASI-LINEAR ELLIPTIC BOUNDARY VALUE PROBLEMS 5.1 Introduction When a physical system depends on more than one variable a general
More information(1:1) 1. The gauge formulation of the Navier-Stokes equation We start with the incompressible Navier-Stokes equation 8 >< >: u t +(u r)u + rp = 1 Re 4
Gauge Finite Element Method for Incompressible Flows Weinan E 1 Courant Institute of Mathematical Sciences New York, NY 10012 Jian-Guo Liu 2 Temple University Philadelphia, PA 19122 Abstract: We present
More informationarxiv: v1 [math.na] 21 Oct 2014
Computing Symmetric Positive Definite Solutions of Three Types of Nonlinear Matrix Equations arxiv:1410.5559v1 [math.na] 21 Oct 2014 Negin Bagherpour a, Nezam Mahdavi-Amiri a, a Department of Mathematical
More information3.1 Interpolation and the Lagrange Polynomial
MATH 4073 Chapter 3 Interpolation and Polynomial Approximation Fall 2003 1 Consider a sample x x 0 x 1 x n y y 0 y 1 y n. Can we get a function out of discrete data above that gives a reasonable estimate
More informationA Comparison of Adomian and Generalized Adomian Methods in Solving the Nonlinear Problem of Flow in Convergent-Divergent Channels
Applied Mathematical Sciences, Vol. 8, 2014, no. 7, 321-336 HIKARI Ltd, www.m-hikari.com http://dx.doi.org/10.12988/ams.2014.39495 A Comparison of Adomian and Generalized Adomian Methods in Solving the
More informationNumerical Methods - Boundary Value Problems for ODEs
Numerical Methods - Boundary Value Problems for ODEs Y. K. Goh Universiti Tunku Abdul Rahman 2013 Y. K. Goh (UTAR) Numerical Methods - Boundary Value Problems for ODEs 2013 1 / 14 Outline 1 Boundary Value
More informationCS 257: Numerical Methods
CS 57: Numerical Methods Final Exam Study Guide Version 1.00 Created by Charles Feng http://www.fenguin.net CS 57: Numerical Methods Final Exam Study Guide 1 Contents 1 Introductory Matter 3 1.1 Calculus
More informationACCURATE NUMERICAL METHOD FOR BLASIUS PROBLEM FOR FLOW PAST A FLAT PLATE WITH MASS TRANSFER
FDS-2000 Conference, pp. 1 10 M. Sapagovas et al. (Eds) c 2000 MII ACCURATE UMERICAL METHOD FOR BLASIUS PROBLEM FOR FLOW PAST A FLAT PLATE WITH MASS TRASFER B. GAHA 1, J. J. H. MILLER 2, and G. I. SHISHKI
More informationB-splines Collocation Algorithms for Solving Numerically the MRLW Equation
ISSN 1749-889 (print), 1749-897 (online) International Journal of Nonlinear Science Vol.8(2009) No.2,pp.11-140 B-splines Collocation Algorithms for Solving Numerically the MRLW Equation Saleh M. Hassan,
More informationHOMOTOPY ANALYSIS METHOD FOR SOLVING KDV EQUATIONS
Surveys in Mathematics and its Applications ISSN 1842-6298 (electronic), 1843-7265 (print) Volume 5 (21), 89 98 HOMOTOPY ANALYSIS METHOD FOR SOLVING KDV EQUATIONS Hossein Jafari and M. A. Firoozjaee Abstract.
More informationIntroductory Numerical Analysis
Introductory Numerical Analysis Lecture Notes December 16, 017 Contents 1 Introduction to 1 11 Floating Point Numbers 1 1 Computational Errors 13 Algorithm 3 14 Calculus Review 3 Root Finding 5 1 Bisection
More informationBoundary Value Problems and Approximate Solutions
Boundary Value Problems and Approximate Solutions Gebreslassie Tesfy, Venketeswara Rao J*, Ataklti Araya and Daniel Tesfay Department of Mathematics, College of Natural and Computational Scineces, Mekelle
More informationExample 2: a system of coupled ODEs with algebraic property at infinity
Example 2: a system of coupled ODEs with algebraic property at infinity Consider a set of two coupled nonlinear differential equations 5 subject to f (η) + θ(η) f 2 = 0, (10) θ (η) = 3σf (η)θ(η), (11)
More informationReview. Numerical Methods Lecture 22. Prof. Jinbo Bi CSE, UConn
Review Taylor Series and Error Analysis Roots of Equations Linear Algebraic Equations Optimization Numerical Differentiation and Integration Ordinary Differential Equations Partial Differential Equations
More informationAdaptive Time Space Discretization for Combustion Problems
Konrad-Zuse-Zentrum fu r Informationstechnik Berlin Takustraße 7 D-14195 Berlin-Dahlem Germany JENS LANG 1,BODO ERDMANN,RAINER ROITZSCH Adaptive Time Space Discretization for Combustion Problems 1 Talk
More informationTHERE are several types of non-newtonian fluid models
INTERNATIONAL JOURNAL OF MATHEMATICS AND SCIENTIFIC COMPUTING (ISSN: 221-5), VOL. 4, NO. 2, 214 74 Invariance Analysis of Williamson Model Using the Method of Satisfaction of Asymptotic Boundary Conditions
More informationNatural Convection Flow with Combined Buoyancy Effects Due to Thermal and Mass Diffusion in a Thermally Stratified Media
Nonlinear Analysis: Modelling and Control, 2004, Vol. 9, No., 89 02 Natural Convection Flow with Combined Buoyancy Effects Due to Thermal and Mass Diffusion in a Thermally Stratified Media Received: 09.2.2003
More informationCOMBINED EFFECTS OF RADIATION AND JOULE HEATING WITH VISCOUS DISSIPATION ON MAGNETOHYDRODYNAMIC FREE CONVECTION FLOW AROUND A SPHERE
Suranaree J. Sci. Technol. Vol. 20 No. 4; October - December 2013 257 COMBINED EFFECTS OF RADIATION AND JOULE HEATING WITH VISCOUS DISSIPATION ON MAGNETOHYDRODYNAMIC FREE CONVECTION FLOW AROUND A SPHERE
More informationResearch Article Lie Group Analysis of Unsteady Flow and Heat Transfer over a Porous Surface for a Viscous Fluid
Journal of Applied Mathematics Volume 202, Article ID 675287, 7 pages doi:0.55/202/675287 Research Article Lie Group Analysis of Unsteady Flow and Heat Transfer over a Porous Surface for a Viscous Fluid
More informationMultiscale Hydrodynamic Phenomena
M2, Fluid mechanics 2014/2015 Friday, December 5th, 2014 Multiscale Hydrodynamic Phenomena Part I. : 90 minutes, NO documents 1. Quick Questions In few words : 1.1 What is dominant balance? 1.2 What is
More informationAdditional exercises with Numerieke Analyse
Additional exercises with Numerieke Analyse March 10, 017 1. (a) Given different points x 0, x 1, x [a, b] and scalars y 0, y 1, y, z 1, show that there exists at most one polynomial p P 3 with p(x i )
More informationJournal of Engineering Science and Technology Review 2 (1) (2009) Research Article
Journal of Engineering Science and Technology Review 2 (1) (2009) 118-122 Research Article JOURNAL OF Engineering Science and Technology Review www.jestr.org Thin film flow of non-newtonian fluids on a
More informationMB4018 Differential equations
MB4018 Differential equations Part II http://www.staff.ul.ie/natalia/mb4018.html Prof. Natalia Kopteva Spring 2015 MB4018 (Spring 2015) Differential equations Part II 0 / 69 Section 1 Second-Order Linear
More informationDEDUCTIVE group transformation analysis, also called
INTERNATIONAL JOURNAL OF MATHEMATICS AND SCIENTIFIC COMPUTING ISSN: 2231-5330), VOL. 2, NO. 2, 2012 53 Deductive Group Invariance Analysis of Boundary Layer Equations of a Special Non-Newtonian Fluid over
More informationMULTISTAGE HOMOTOPY ANALYSIS METHOD FOR SOLVING NON- LINEAR RICCATI DIFFERENTIAL EQUATIONS
MULTISTAGE HOMOTOPY ANALYSIS METHOD FOR SOLVING NON- LINEAR RICCATI DIFFERENTIAL EQUATIONS Hossein Jafari & M. A. Firoozjaee Young Researchers club, Islamic Azad University, Jouybar Branch, Jouybar, Iran
More informationScientific Computing: An Introductory Survey
Scientific Computing: An Introductory Survey Chapter 7 Interpolation Prof. Michael T. Heath Department of Computer Science University of Illinois at Urbana-Champaign Copyright c 2002. Reproduction permitted
More informationVariation of Parameters Method for Solving Fifth-Order. Boundary Value Problems
Applied Mathematics & Information Sciences 2(2) (28), 135 141 An International Journal c 28 Dixie W Publishing Corporation, U. S. A. Variation of Parameters Method for Solving Fifth-Order Boundary Value
More informationInternational Journal of Mathematical Archive-8(1), 2017, Available online through ISSN
International Journal of Mathematical Archive-8(), 7, 46-55 Available online through.ijma.info ISSN 9 546 MHD BOUNDARY LAYER SLIP FLOW AND HEAT TRANSFER OVER A POROUS FLAT PLATE EMBEDDED IN A POROUS MEDIUM
More information(Received 11 October 1996; accepted 7 March 1997)
Padé approximant algorithm for solving nonlinear ordinary differential equation boundary value problems on an unbounded domain John P. Boyd a Department of Atmospheric, Oceanic and Space Science, University
More informationA Neural Network Study of Blasius Equation
A Neural Network Study of Blasius Equation Halil Mutuk Physics Department, Ondokuz Mayis University, Samsun, Turkey In this work we applied a feed forward neural network to solve Blasius equation which
More informationNumerical Methods for Engineers. and Scientists. Applications using MATLAB. An Introduction with. Vish- Subramaniam. Third Edition. Amos Gilat.
Numerical Methods for Engineers An Introduction with and Scientists Applications using MATLAB Third Edition Amos Gilat Vish- Subramaniam Department of Mechanical Engineering The Ohio State University Wiley
More informationAn Efficient Algorithm Based on Quadratic Spline Collocation and Finite Difference Methods for Parabolic Partial Differential Equations.
An Efficient Algorithm Based on Quadratic Spline Collocation and Finite Difference Methods for Parabolic Partial Differential Equations by Tong Chen A thesis submitted in conformity with the requirements
More informationOrdinary Differential Equations
CHAPTER 8 Ordinary Differential Equations 8.1. Introduction My section 8.1 will cover the material in sections 8.1 and 8.2 in the book. Read the book sections on your own. I don t like the order of things
More informationHydromagnetic Flow Near a Stagnation Point on a Stretching Sheet with Variable Thermal Conductivity and Heat Source/Sink
International Journal of Applied Science and Engineering 2013. 11, 3: 331-341 Hydromagnetic Flow Near a Stagnation Point on a Stretching Sheet with Variable Thermal Conductivity and Heat Source/Sink J.
More informationExplicit Solution of Axisymmetric Stagnation. Flow towards a Shrinking Sheet by DTM-Padé
Applied Mathematical Sciences, Vol. 4, 2, no. 53, 267-2632 Explicit Solution of Axisymmetric Stagnation Flow towards a Shrining Sheet by DTM-Padé Mohammad Mehdi Rashidi * Engineering Faculty of Bu-Ali
More informationApplied Numerical Analysis
Applied Numerical Analysis Using MATLAB Second Edition Laurene V. Fausett Texas A&M University-Commerce PEARSON Prentice Hall Upper Saddle River, NJ 07458 Contents Preface xi 1 Foundations 1 1.1 Introductory
More informationFlow and Natural Convection Heat Transfer in a Power Law Fluid Past a Vertical Plate with Heat Generation
ISSN 1749-3889 (print), 1749-3897 (online) International Journal of Nonlinear Science Vol.7(2009) No.1,pp.50-56 Flow and Natural Convection Heat Transfer in a Power Law Fluid Past a Vertical Plate with
More informationAnalytical Solution of BVPs for Fourth-order Integro-differential Equations by Using Homotopy Analysis Method
ISSN 1749-3889 (print), 1749-3897 (online) International Journal of Nonlinear Science Vol.9(21) No.4,pp.414-421 Analytical Solution of BVPs for Fourth-order Integro-differential Equations by Using Homotopy
More informationON THE EFFECTIVENESS OF HEAT GENERATION/ABSORPTION ON HEAT TRANSFER IN A STAGNATION POINT FLOW OF A MICROPOLAR FLUID OVER A STRETCHING SURFACE
5 Kragujevac J. Sci. 3 (29) 5-9. UDC 532.5:536.24 ON THE EFFECTIVENESS OF HEAT GENERATION/ABSORPTION ON HEAT TRANSFER IN A STAGNATION POINT FLOW OF A MICROPOLAR FLUID OVER A STRETCHING SURFACE Hazem A.
More informationA derivative-free nonmonotone line search and its application to the spectral residual method
IMA Journal of Numerical Analysis (2009) 29, 814 825 doi:10.1093/imanum/drn019 Advance Access publication on November 14, 2008 A derivative-free nonmonotone line search and its application to the spectral
More informationCommun Nonlinear Sci Numer Simulat
Commun Nonlinear Sci Numer Simulat 15 (21) 3965 3973 Contents lists available at ScienceDirect Commun Nonlinear Sci Numer Simulat journal homepage: www.elsevier.com/locate/cnsns A similarity solution in
More informationIn practice one often meets a situation where the function of interest, f(x), is only represented by a discrete set of tabulated points,
1 Interpolation 11 Introduction In practice one often meets a situation where the function of interest, f(x), is only represented by a discrete set of tabulated points, {x i, y i = f(x i ) i = 1 n, obtained,
More informationINTERPOLATION. and y i = cos x i, i = 0, 1, 2 This gives us the three points. Now find a quadratic polynomial. p(x) = a 0 + a 1 x + a 2 x 2.
INTERPOLATION Interpolation is a process of finding a formula (often a polynomial) whose graph will pass through a given set of points (x, y). As an example, consider defining and x 0 = 0, x 1 = π/4, x
More informationAnumerical and analytical study of the free convection thermal boundary layer or wall jet at
REVERSED FLOW CALCULATIONS OF HIGH PRANDTL NUMBER THERMAL BOUNDARY LAYER SEPARATION 1 J. T. Ratnanather and P. G. Daniels Department of Mathematics City University London, England, EC1V HB ABSTRACT Anumerical
More informationQUINTIC SPLINE SOLUTIONS OF FOURTH ORDER BOUNDARY-VALUE PROBLEMS
INTERNATIONAL JOURNAL OF NUMERICAL ANALYSIS AND MODELING Volume 5, Number, Pages 0 c 2008 Institute for Scientific Computing Information QUINTIC SPLINE SOLUTIONS OF FOURTH ORDER BOUNDARY-VALUE PROBLEMS
More informationExam 2. Average: 85.6 Median: 87.0 Maximum: Minimum: 55.0 Standard Deviation: Numerical Methods Fall 2011 Lecture 20
Exam 2 Average: 85.6 Median: 87.0 Maximum: 100.0 Minimum: 55.0 Standard Deviation: 10.42 Fall 2011 1 Today s class Multiple Variable Linear Regression Polynomial Interpolation Lagrange Interpolation Newton
More informationHierarchical Reconstruction with up to Second Degree Remainder for Solving Nonlinear Conservation Laws
Hierarchical Reconstruction with up to Second Degree Remainder for Solving Nonlinear Conservation Laws Dedicated to Todd F. Dupont on the occasion of his 65th birthday Yingjie Liu, Chi-Wang Shu and Zhiliang
More informationQuarter-Sweep Gauss-Seidel Method for Solving First Order Linear Fredholm Integro-differential Equations
MATEMATIKA, 2011, Volume 27, Number 2, 199 208 c Department of Mathematical Sciences, UTM Quarter-Sweep Gauss-Seidel Method for Solving First Order Linear Fredholm Integro-differential Equations 1 E. Aruchunan
More informationA SEMI-ANALYTICAL ANALYSIS OF A FREE CONVECTION BOUNDARY-LAYER FLOW OVER A VERTICAL PLATE
A SEMI-ANALYTICAL ANALYSIS OF A FREE CONVECTION BOUNDARY-LAYER FLOW OVER A VERTICAL PLATE Haldun Alpaslan PEKER and Galip OTURANÇ Department of Mathematics, Faculty of Science, Selcu University, 475, Konya,
More informationAn Efficient Multiscale Runge-Kutta Galerkin Method for Generalized Burgers-Huxley Equation
Applied Mathematical Sciences, Vol. 11, 2017, no. 30, 1467-1479 HIKARI Ltd, www.m-hikari.com https://doi.org/10.12988/ams.2017.7141 An Efficient Multiscale Runge-Kutta Galerkin Method for Generalized Burgers-Huxley
More informationCS 323: Numerical Analysis and Computing
CS 323: Numerical Analysis and Computing MIDTERM #2 Instructions: This is an open notes exam, i.e., you are allowed to consult any textbook, your class notes, homeworks, or any of the handouts from us.
More informationEXISTENCE OF SOLUTIONS FOR KIRCHHOFF TYPE EQUATIONS WITH UNBOUNDED POTENTIAL. 1. Introduction In this article, we consider the Kirchhoff type problem
Electronic Journal of Differential Equations, Vol. 207 (207), No. 84, pp. 2. ISSN: 072-669. URL: http://ejde.math.txstate.edu or http://ejde.math.unt.edu EXISTENCE OF SOLUTIONS FOR KIRCHHOFF TYPE EQUATIONS
More informationA general theory of discrete ltering. for LES in complex geometry. By Oleg V. Vasilyev AND Thomas S. Lund
Center for Turbulence Research Annual Research Briefs 997 67 A general theory of discrete ltering for ES in complex geometry By Oleg V. Vasilyev AND Thomas S. und. Motivation and objectives In large eddy
More informationCommun Nonlinear Sci Numer Simulat
Commun Nonlinear Sci Numer Simulat 16 (2011) 2730 2736 Contents lists available at ScienceDirect Commun Nonlinear Sci Numer Simulat journal homepage: www.elsevier.com/locate/cnsns Homotopy analysis method
More informationWe consider the problem of finding a polynomial that interpolates a given set of values:
Chapter 5 Interpolation 5. Polynomial Interpolation We consider the problem of finding a polynomial that interpolates a given set of values: x x 0 x... x n y y 0 y... y n where the x i are all distinct.
More informationLAMINAR BOUNDARY LAYER FLOW OVER A CONE WITH SUCTION OR INJECTION*
QUARTERLY OF APPLIED MATHEMATICS VOLUME XLVI, NUMBER 1 MARCH 1988, PAGES 145-156 LAMINAR BOUNDARY LAYER FLOW OVER A CONE WITH SUCTION OR INJECTION* By TAKASHI WATANABE Iwate University, Morioka, Japan
More informationOrdinary Differential Equations
Ordinary Differential Equations We call Ordinary Differential Equation (ODE) of nth order in the variable x, a relation of the kind: where L is an operator. If it is a linear operator, we call the equation
More informationNonlinear Analysis: Modelling and Control, 2008, Vol. 13, No. 4,
Nonlinear Analysis: Modelling and Control, 2008, Vol. 13, No. 4, 513 524 Effects of Temperature Dependent Thermal Conductivity on Magnetohydrodynamic (MHD) Free Convection Flow along a Vertical Flat Plate
More informationPreliminary Examination in Numerical Analysis
Department of Applied Mathematics Preliminary Examination in Numerical Analysis August 7, 06, 0 am pm. Submit solutions to four (and no more) of the following six problems. Show all your work, and justify
More informationSolving Linear Sixth-Order Boundary Value Problems by Using Hyperbolic Uniform Spline Method
International Journal of Mathematical Modelling & Computations Vol. 03, No. 03, 013, 169-180 Solving Linear Sixth-Order Boundary Value Problems by Using Hyperbolic Uniform Spline Method J. Dabounou a,
More informationA Numerical Solution of the Lax s 7 th -order KdV Equation by Pseudospectral Method and Darvishi s Preconditioning
Int. J. Contemp. Math. Sciences, Vol. 2, 2007, no. 22, 1097-1106 A Numerical Solution of the Lax s 7 th -order KdV Equation by Pseudospectral Method and Darvishi s Preconditioning M. T. Darvishi a,, S.
More informationMA22S3 Summary Sheet: Ordinary Differential Equations
MA22S3 Summary Sheet: Ordinary Differential Equations December 14, 2017 Kreyszig s textbook is a suitable guide for this part of the module. Contents 1 Terminology 1 2 First order separable 2 2.1 Separable
More informationIntroduction to the Numerical Solution of IVP for ODE
Introduction to the Numerical Solution of IVP for ODE 45 Introduction to the Numerical Solution of IVP for ODE Consider the IVP: DE x = f(t, x), IC x(a) = x a. For simplicity, we will assume here that
More informationDifferential Equations
Differential Equations Definitions Finite Differences Taylor Series based Methods: Euler Method Runge-Kutta Methods Improved Euler, Midpoint methods Runge Kutta (2nd, 4th order) methods Predictor-Corrector
More informationSimilarity Approach to the Problem of Second Grade Fluid Flows over a Stretching Sheet
Applied Mathematical Sciences, Vol. 1, 2007, no. 7, 327-338 Similarity Approach to the Problem of Second Grade Fluid Flows over a Stretching Sheet Ch. Mamaloukas Athens University of Economics and Business
More informationMultilevel Preconditioning of Graph-Laplacians: Polynomial Approximation of the Pivot Blocks Inverses
Multilevel Preconditioning of Graph-Laplacians: Polynomial Approximation of the Pivot Blocks Inverses P. Boyanova 1, I. Georgiev 34, S. Margenov, L. Zikatanov 5 1 Uppsala University, Box 337, 751 05 Uppsala,
More informationSIMILARITY SOLUTION FOR MHD FLOW THROUGH VERTICAL POROUS PLATE WITH SUCTION
Journal of Computational and Applied Mechanics, Vol. 6., No. 1., (2005), pp. 15 25 SIMILARITY SOLUTION FOR MHD FLOW THROUGH VERTICAL POROUS PLATE WITH SUCTION Mohammad Ferdows, Masahiro Ota Department
More informationThe Riccati transformation method for linear two point boundary problems
Chapter The Riccati transformation method for linear two point boundary problems The solution algorithm for two point boundary value problems to be employed here has been derived from different points
More information