Skip to main content
Library homepage
 

Text Color

Text Size

 

Margin Size

 

Font Type

Enable Dyslexic Font
Physics LibreTexts

2.14: Eigenvectors and Eigenvalues

( \newcommand{\kernel}{\mathrm{null}\,}\)

In Sections 11-13, we have been considering some aspects of the moments of inertia of plane laminas, and we have discussed such matters as rotation of axes, and such concepts as product moments of inertia, principal axes, principal moments of inertia and the momental ellipse. We next need to develop the same concepts with respect to three-dimensional solid bodies. In doing so, we shall need to make use of the algebraic concepts of eigenvectors and eigenvalues. If you are already familiar with such matters, you may want to skip this section and move on to the next. If the ideas of eigenvalues and eigenvectors are new to you, or if you are a bit rusty with them, this section may be helpful. I do assume that the reader is at least familiar with the elementary rules of matrix multiplication.

Example 2.14.1

Consider what happens when you multiply a vector, for example the vector (01),

by a square matrix, for example the matrix (4121) we obtain:

(4121)(01)=(11)

The result of the operation is another vector that is in quite a different direction from the original one.

However, now let us multiply the vector (11) by the same matrix. The result is (33). The result of the multiplication is merely to multiply the vector by 3 without changing its direction. The vector (11) is a very special one, and it is called an eigenvector of the matrix, and the multiplier 3 is called the corresponding eigenvalue. "Eigen" is German for "own" in the sense of "my own book". There is one other eigenvector of the matrix; it is the vector (12). Try it; you should that the corresponding eigenvalue is 2.

In short, given a square matrix A, if you can find a vector a such that Aa = λ a, where λ is merely a scalar multiplier that does not change the direction of the vector a, then a is an eigenvector and λ is the corresponding eigenvalue.

In the above, I told you what the two eigenvectors were, and you were able to verify that they were indeed eigenvectors and you were able to find their eigenvalues by straightforward arithmetic. But, what if I hadn't told you the eigenvectors? How would you find them?

Let A=(A11A12A21A22) and let x=(x1x2) be an eigenvector with corresponding eigenvalue λ. Then we must have

(A11A12A21A22)(x1x2)=(λx1λx2).

That is,

(A11λ)x1+A12x2=0

and

A21x1+(A22λ)x2=0.

These two equations are consistent only if the determinant of the coefficients is zero. That is,

[A11λA12A21A22λ] = 0

This equation is a quadratic equation in λ, known as the characteristic equation, and its two roots, the characteristic or latent roots are the eigenvalues of the matrix. Once the eigenvalues are found the ratio of x1 to x2 is easily found, and hence the eigenvectors.

Similarly, if A is a 3 × 3 matrix, the characteristic equation is

[A11λA12A13A21A22λA23A31A32A33λ] = 0

This is a cubic equation in λ, the three roots being the eigenvalues. For each eigenvalue, the ratio x1:x2:x3 can easily be found and hence the eigenvectors. The characteristic equation is a cubic equation, and is best solved numerically, not by algebraic formula. The cubic equation can be written in the form

λ3+a2λ2+a1λ+a0=0,

and the solutions can be checked from the following results from the theory of equations:

λ1+λ2+λ3=a2,

λ2λ3+λ3λ1+λ1λ2=a1,

λ1λ2λ3=a0.


This page titled 2.14: Eigenvectors and Eigenvalues is shared under a CC BY-NC 4.0 license and was authored, remixed, and/or curated by Jeremy Tatum via source content that was edited to the style and standards of the LibreTexts platform.

Support Center

How can we help?