Scribd Logo
Upload

Welcome to Scribd!

  • 28 views

    Ruturaj Arvind Umaranikar 200979009 Assignment 3

    Uploaded by

    Ruturaj Umaranikar
    This document provides the numerical solution for two second order ODEs using boundary element methods. For the first ODE (νu ́ ́ + u ́ + u = 0), the domain is discretized and central differencing is used to solve the equation iteratively using Thomas algorithm. For the non-linear second ODE (u ́ ́ - 2uu ́ = 0), Newton-Raphson method is used to linearize before applying central differencing and Thomas algorithm. The document also provides the numerical solution procedure for the Blasius equation and calculates boundary layer properties for a flat plate.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    Ruturaj Arvind Umaranikar 200979009 Assignment 3

    Uploaded by

    Ruturaj Umaranikar
    28 views13 pages
    This document provides the numerical solution for two second order ODEs using boundary element methods. For the first ODE (νu ́ ́ + u ́ + u = 0), the domain is discretized and central differencing is used to solve the equation iteratively using Thomas algorithm. For the non-linear second ODE (u ́ ́ - 2uu ́ = 0), Newton-Raphson method is used to linearize before applying central differencing and Thomas algorithm. The document also provides the numerical solution procedure for the Blasius equation and calculates boundary layer properties for a flat plate.

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document provides the numerical solution for two second order ODEs using boundary element methods. For the first ODE (νu ́ ́ + u ́ + u = 0), the domain is discretized and central differencing is used to solve the equation iteratively using Thomas algorithm. For the non-linear second ODE (u ́ ́ - 2uu ́ = 0), Newton-Raphson method is used to linearize before applying central differencing and Thomas algorithm. The document also provides the numerical solution procedure for the Blasius equation and calculates boundary layer properties for a flat plate.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    28 views13 pages

    Ruturaj Arvind Umaranikar 200979009 Assignment 3

    Uploaded by

    Ruturaj Umaranikar
    This document provides the numerical solution for two second order ODEs using boundary element methods. For the first ODE (νu ́ ́ + u ́ + u = 0), the domain is discretized and central differencing is used to solve the equation iteratively using Thomas algorithm. For the non-linear second ODE (u ́ ́ - 2uu ́ = 0), Newton-Raphson method is used to linearize before applying central differencing and Thomas algorithm. The document also provides the numerical solution procedure for the Blasius equation and calculates boundary layer properties for a flat plate.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 13

    Advanced Fluid Dynamics

    Assignment No. 3
    Name- Ruturaj Arvind Umaranikar
    Reg. No. - 200979009

    Que. Obtain numerical solution for the second order ODE νu´´ + u´ + u = 0 using boundary
    conditions; u(0) = 0 ; u(1) = 0.5

    Algorithm –

    • Discretize the given domain 0 ≤ η ≤ 1 into J-1 panels and assign each node with the indices
    j = 1, 2, 3,.., J. Here J = 101.

    • Set the kinematic viscosity (ν) value appropriately.


    • Calculate the grid spacing Δη = 1 /(J−1)

    • Apply the central difference algorithm to obtain u´´ and u´ and substitute the values in the
    given ODE.

    Also can be arranged in the form,

    • Calculate a, b, c by using the following relations,

    • By using Thomas Algorithm i.e. Gauss Elimination Method back substitution process we find 𝑢𝑗 by the
    following formula,

    In which,

    Here 𝑗 = 𝑎𝑙𝑝ℎ𝑎 (𝑗) and 𝑗 = 𝑏𝑒𝑡𝑎 (𝑗)

    • Calculated values are plotted in the graph Repeat the process for different values of Kinematic
    Viscosity ()
    Flow Chart-

    Matlab Code-
    clc; clear all;
    v = 0.10;
    J = 101;
    deta = 1 /(J-1);
    a = 1-(deta/(2*v));
    b = -2+((deta^2)/v);
    c = 1+ (deta/(2*v));
    u = 1:J;
    u(1) = 0;
    u(J) = 0.5;
    alpha = 1:J-1;
    beta = 1:J-1;
    alpha(1) = 0;
    beta(1) = 0;
    for k = 1:1:J
    eta (k) = (k-1)*deta;
    end
    for i = 2:1:J-1
    alpha (i) = -c/((a*alpha(i-1))+b);
    beta (i) = ((-a)*beta(i-1))/(a*alpha(i-1)+b);
    end
    for j = J-1:-1:2
    u(j) = alpha(j)*u(j+1)+ beta(j);
    end
    plot (eta, u, 'b')
    hold on
    %plot (eta, u, 'g')
    hold on
    %plot (eta, u, 'r')
    hold on

    Plot-

    Que- Obtain numerical solution for the non-linear ODE u´´ - 2uu´ = 0 using boundary conditions;
    u(0) = 0, u(5) = 1.

    Algorithm

    • Discretize the given domain 0 ≤ η ≤ 5 into J-1 panels, and assign each node with the indices
    j = 1, 2, 3,..., J. Here J = 101.
    • Calculate the grid spacing Δη = 5/( 𝐽−1).
    • Using Taylor series and Newton Raphson method to discretize the non-linear part of the ODE, we
    get
    • Applying the central difference algorithm we get following,

    Or can be arranged as follows,

    Where,

    • The term 𝑢𝑗 and 𝑢𝑗 ′can be initialized using an appropriate function, Here 𝑢𝑗 = 𝑈.

    • Apply the Thomas algorithm to obtain 𝑢𝑗

    Where,

    With boundary conditions,


    Here 𝑗 = 𝑎𝑙𝑝ℎ𝑎 (𝑗) and 𝑗 = 𝑏𝑒𝑡𝑎 (𝑗)

    • Compare the values of calculated u and initialized u. If they are the same then stop iteration,
    else continue till they are the same

    • Plot the results.

    Flow Chart-
    Matlab Code

    clc;

    clear all;
    J = 101;
    deleta = 5/(J-1);
    U = 1:J;
    U1 = 1:J;
    U2 = 1:J;
    A = 1:J;
    B = 1:J;
    C = 1:J;
    D = 1:J;
    alpha = 1:J;
    beta = 1:J;
    U(1) = 0 ; U(J) = 1;
    alpha(1) = 0 ; beta(1) = 0;
    e = 0.1;
    for n = 1:1:J;
    eta(n)= (n-1)*deleta;
    U(n)= sqrt (eta(n)/5);
    if 0<eta(n)<1;
    U1(n)=eta(n);
    else
    U1(n)=1;
    end
    end
    for j= 2 :1 :J;
    for m = 2:1:(J-1);
    U2(m)= ((U1(m+1)-U1(m-1))/(2*deleta));
    A(m)= 1+(deleta*U1(m));
    B(m)= -2*(1+((deleta^2)*U2(m)));
    C(m)= 1-(deleta*U1(m));
    D(m)= (-2)*(deleta^2)*U1(m)*U2(m);
    alpha(m)= ((-C(m))/((A(m)*alpha(m-1))+B(m)));
    beta(m)= (D(m)-(A(m)*beta(m-1)))/((A(m)*alpha(m-1))+B(m));
    end
    for k = (J-1):-1:2
    U(k)= (alpha(k)*U(k+1))+beta(k);
    end
    error = abs (U(j)-U1(j))
    if error < e
    U(j)=U(j);
    else
    U1(j)=U(j);
    end
    end
    plot ( eta, U )
    hold on

    Plot-

    𝟏
    Que. Obtain numerical solution of Blasius equation f´´´ + 𝟐 f f´´ = 0 under the boundary conditions

    f (0) = f ´ (0) = 0, f ´(ꝏ) = 1

    Algorithm

    • Divide the given domain 0 ≤ η ≤ 10 into J-1 panels, and assign each node with the indices
    j = 1, 2, 3,..., J. Here J = 101.
    10
    • Calculate the grid spacing Δη = 𝐽−1

    • For efficient formulation, a new variable g defined by

    Setting of initial values of 𝑓𝑗 and 𝑔𝑗 . In coding 𝑓𝑗 is F and 𝑔𝑗 is G and 𝑔′𝑗 is GD.


    • Using central difference formula, calculate values of f ´, g ´, g ´´
    • Calculate the values of 𝐴𝑗 , 𝐵𝑗 , 𝐶𝑗 , 𝐷𝑗 , 𝐸𝑗 .

    • Use Thomas algorithm to calculate 𝑔𝑗 . For that first calculating the values of 𝑅𝑗 , 𝑆𝑗 , 𝑇𝑗 , 𝑉𝑗 .

    𝐽−1
    • Apply central difference formula at a point to calculate 𝑓𝑗 .
    2
    • Compare the values of calculated f, g and initialised F, G. If they are same then stop iteration,
    else continue till they are same.

    • Plot the results.


    Flow Chart-
    Matlab Code -

    clc;
    clear all;

    J = 101; n0 = 0; n1 = 10;
    deta = (n1-n0)/(J-1);
    e = 0.01;
    f = 1:J; g = 1:J;
    eta = 1:J; GD = 1:J;
    f(1) = 0; g(1) = 0;
    G = 1:J;
    F = 1:J;
    A = 1:J; B = 1:J;
    C = 1:J; D = 1:J; E = 1:J;
    R = 1:J; S = 1:J; T = 1:J; V =1:J;
    R(J) = 1; S(J) = 0; T(J) = 0;
    for i = 1:1:J
    eta (i) = (i-1)*deta;
    end
    for m = 1:1:J
    if 0<eta(m)<1
    F (m) = ((eta(m))^2)/2;
    G (m) = eta (m);
    else
    F (m) = eta(m)-(1/2);
    G (m) = 1;
    end
    end
    for n= 2:1:J
    for k = 2:1:J-1
    GD (k) = (G(k+1)-G(k-1))/(2*deta);
    A (k) = 1+((deta*F(k))/4);
    B (k) = -2;
    C (k) = 1-((deta*F(k))/4);
    D (k) =(deta^2)*GD(k)/2;
    E (k) = (deta^2)*F(k)*GD(k)/2;
    end
    for z = J-1:-1:2
    V(z) = (A(z)*S(z+1))+B(z)+(deta*(A(z)*T(z+1))+D(z))/2;
    R(z) = (E(z)-(A(z)*R(z+1)))/V(z);
    S(z) = -(C(z)+(deta*(A(z)*T(z+1)+D(z))/2))/V(z);
    T(z) = -(A(z)*T(z+1)+D(z))/V(z);
    end

    for j = 2:1:J
    g(j)= R(j)+S(j)*g(j-1)+T(j)*f(j-1);
    f(j)= f(j-1)+(deta/2)*(g(j)+g(j-1));
    end
    error = abs (g(n)-G(n))
    if error < e
    f(n)=f(n);
    g(n)=g(n);
    else
    F(n) = f(n);
    G(n) = g(n);
    end
    end
    plot (eta,g,'g')
    hold on
    plot (eta,f,'r')

    Plot -
    Analysis of Flat Plate Boundary Layer-

    Let’s select Reynold’s Number Re = 3 × 105 and as X is a dimensionless quantity so it varies only
    between 0 to 1. We are analysing at mid span of plate length and at its trailing edge so we select
    X = 0.5 and X = 1, 𝐿∗ is the length of plate in m and is 8.5 m.
    ∗
    1. Boundary Layer Thickness - 𝛿99 , 𝑚.
    5 × √𝑋
    𝛿99 =
    √𝑅𝑒

    ∗
    𝛿99 = 𝛿99 × 𝐿∗
    2. Displacement Thickness - 𝛿1∗ , 𝑚.
    1.7208 × √𝑋
    𝛿1 =
    √𝑅𝑒

    𝛿1∗ = 𝛿1 × 𝐿∗
    3. Momentum Thickness – 𝛿2∗ , 𝑚.
    0.664 × √𝑋
    𝛿2 =
    √𝑅𝑒

    𝛿2∗ = 𝛿2 × 𝐿∗

    ∗
    4. For given values, calculations of Boundary Layer Thickness - 𝛿99 , Displacement Thickness -
    𝛿1∗ , Momentum Thickness – 𝛿2∗ are tabulated below,

    X 𝛿99 𝛿1 𝛿2
    0.5 0.006455 0.002222 0.000857
    1 0.009129 0.003142 0.001212

    ∗
    𝛿99 = 𝛿99 × 𝐿∗ 𝛿1∗ = 𝛿1 × 𝐿∗ 𝛿2∗ = 𝛿2 × 𝐿∗
    0.01883 0.007286 0.054867
    0.026705 0.010304 0.077594

    You might also like