Solution of Matrix Eigenvalue Problem

Transcription

1 Outlines October 12, 2004

2 Outlines Part I: Review of Previous Lecture Part II: Review of Previous Lecture

3 Outlines Part I: Review of Previous Lecture Part II: Standard Matrix Eigenvalue Problem Other Forms of the Standard Eigenvalue Problem Solvability of the Standard Eigenvalue Problem Example Standard Eigenvalue Problems s

4 Part I Review of Previous Lecture

5 Review of Previous Lecture Jacobi Iteration Method Gauss-Seidel Iteration Method Use of Software Packages

6 Part II

7 Other Forms of the Standard Eigenvalue Problem Solvability of the Standard Eigenvalue Problem Standard Matrix Eigenvalue Problem: Consider a system of equations in algebraic form (a 11 λ)x 1 + a 12 x 2 + a 13 x a 1n x n =0 a 21 x 1 + (a 22 λ)x 2 + a 23 x a 2n x n =0. a n1 x 1 + a n2 x 2 + a n3 x (a nn λ)x n =0 This is not a normal system of linear algebraic equations we re used to. For one, there are n equations, but n + 1 unknowns (the x i values, and also λ). This particular system of equations is known as the standard eigenvalue problem.

8 Other Forms of the Standard Eigenvalue Problem Solvability of the Standard Eigenvalue Problem Form 1 The three forms shown are all algebraically equivalent. Any system of equations that can be expressed in these forms is a standard eigenvalue problem. a 11 λ a 12 a 1n a 21 a 22 λ a 2n a n1 a n2 a nn λ x 1 x 2. x n =

9 Other Forms of the Standard Eigenvalue Problem Solvability of the Standard Eigenvalue Problem Form 2 a 11 a 12 a 1n a 21 a 22 a 2n a n1 a n2 a nn λ x 1 x 2. x n = ([A] λ [I ]) {x} = {0}

10 Other Forms of the Standard Eigenvalue Problem Solvability of the Standard Eigenvalue Problem Form 3 a 11 a 12 a 1n a 21 a 22 a 2n a n1 a n2 a nn x 1 x 2. x n [A] {x} = λ {x} = λ x 1 x 2. x n

11 Other Forms of the Standard Eigenvalue Problem Solvability of the Standard Eigenvalue Problem Solvability of the Standard Eigenvalue Problem Recall form 2 of the standard eigenvalue problem: ([A] λ [I ]) {x} = {0} This system of equations has a solution for values of λ that cause the determinant of the coefficient matrix to equal 0, that is: [A] λ [I ] = 0

12 Other Forms of the Standard Eigenvalue Problem Solvability of the Standard Eigenvalue Problem Characteristic Equation Expanding out all the terms of the previous determinant a 11 λ a 12 a 1n a 21 a 22 λ a 2n..... = 0. a n1 a n2 a nn λ yields a long polynomial in terms of λ. This polynomial will be nth order, and will therefore have n roots, each of which may be real or complex.

13 Example : Many physical systems do not automatically present themselves as a standard eigenvalue problem, even though they can be reformatted as a standard eigenvalue problem. The form of a general eigenvalue problem is [A] {x} = λ [B] {x} where [A] and [B] are symmetric matrices of size n n.

14 Example Example A forging hammer of mass m 2 is mounted on a concrete foundation block of mass m 1. The stiffnesses of the springs underneath the forging hammer and the foundation block are given by k 2 and k 1, respectively.

15 Example Example The system undergoes simple harmonic motion at one of its natural frequencies ω. That is: x 1 (t) = cos(ωt + φ 1 ) x 2 (t) = cos(ωt + φ 2 ) a 1 (t) = ω 2 x 1 (t) a 2 (t) = ω 2 x 2 (t)

16 Example Example Each mass in the system obeys Newton s second law of motion, that is: ΣF = ma Forces on the foundation block: forces from the lower springs, which counteracts motion in the x direction at an amount k 1 x 1 forces from the upper springs, which act according to the amount of relative displacement of the masses m 1 and m 2 : k 2 (x 1 x 2 )

17 Example Example The equilibrium equation for the foundation mass is then ΣF = ma k 1 x 1 k 2 (x 1 x 2 ) = m 1 a ( k 2 k 1 )x 1 + k 2 x 2 = m 1 a ( k 2 k 1 )x 1 + k 2 x 2 = m 1 ω 2 x 1 (k 1 + k 2 )x 1 k 2 x 2 = m 1 ω 2 x 1

18 Example Example Similarly, the equilibrium equation for the forging hammer mass is k 2 x 1 + k 2 x 2 = m 2 ω 2 x 2

19 Example Example So the two equations of motion are (k 1 + k 2 )x 1 k 2 x 2 = m 1 ω 2 x 1 k 2 x 1 + k 2 x 2 = m 2 ω 2 x 2 or in matrix form [ k1 + k 2 k 2 k 2 k 2 ] { x1 x 2 } = ω 2 [ m1 0 0 m 2 ] { x1 x 2 } This is a general eigenvalue problem [A]{x} = λ[b]{x} where [A] is the spring matrix, {x} is the vector of x values, λ is ω 2, and [B] is the mass matrix.

20 Standard Eigenvalue Problems s : Standard Problems The design of a mechanical component requires that the maximum principal stress to be less than the material strength. For a component subjected to arbitrary loads, the principal stresses σ are given by the solution of the equation σ xx τ xy τ xz τ xy σ yy τ yz τ xz τ yz σ zz l x l y l z = σ where the σ values represent normal stresses in the x, y, and z directions, and the τ values represent shear stresses in the xy, xz, and yz planes. The l values represent direction cosines that define the principal planes on which the principal stress occurs. l x l y l z

21 Standard Eigenvalue Problems s : Standard Problems Determine the principal stresses and principal planes in a machine component for the following stress condition σ xx τ xy τ xz τ xy σ yy τ yz τ xz τ yz σ zz = MPa

22 Standard Eigenvalue Problems s MATLAB Solution clear all sigma =[ ]; [dirs, stresses ]= eig ( sigma ); % diag ( A) extracts the elements of the % [ A] matrix along the diagonal principalstresslist = diag ( stresses ) principaldirs = dirs

23 Standard Eigenvalue Problems s MATLAB Solution >> rao_ p431 principalstresslist = principaldirs =

24 Standard Eigenvalue Problems s : General Problems Solve the forging hammer problem for the following values: m 1 = kg m 2 = 5000 kg k 1 = N/m k 2 = N/m

25 Standard Eigenvalue Problems s Solving Eigenvalue Problems in MATLAB Solving this eigenvalue problem will yield 2 eigenvalues equal to the square of the system s natural frequencies, and 2 corresponding x vector values that show the relative displacements of the m 1 and m 2 masses at those frequencies.

26 Standard Eigenvalue Problems s MATLAB Solution (Part 1) clear all ; % Define spring constants and masses % for hammer and foundation block k1 =1 e7; k2 =5 e6; m1 =20000; m2 =5000; % Define system stiffness matrix K=[ k1+k2 -k2 -k2 k2 ]; % Define system mass matrix M=[ m1 0 0 m2 ];

27 Standard Eigenvalue Problems s MATLAB Solution (Part 2) % Solve general eigenvalue problem [X, Omega2 ]= eig (K,M); % diag ( A) extracts the elements of the % [ A] matrix along the diagonal Omega = diag ( sqrt ( Omega2 )); % Scale column 1 of the [ X] matrix by % the row 1, column 1 X value X (:,1)= X (:,1)/ X (1,1); % Scale column 2 of the [ X] matrix by % the row 1, column 2 X value X (:,2)= X (:,2)/ X (1,2); Omega X

28 Standard Eigenvalue Problems s MATLAB Solution (Results) >> rao_ ex42 Omega = X =

29 Continue working on the Gauss-Seidel and other homework problems already assigned.