0
$\begingroup$

I'm having difficulty with this question from my Linear Algebra course,

Question details

I don't know how to answer the following,

Find the axis of rotation of the composition L1 ◦ L2

$\endgroup$
1
  • $\begingroup$ Have you learned about eigenvalues and eigenvectors yet? $\endgroup$ Apr 25 '18 at 12:44
0
$\begingroup$

What you're looking for is the eigenvector associated with $\lambda = 1$ for the composition $L_1 \circ L_2$. In other words, you want a non-zero vector $v$ which will solve the equation $L_1(L_2(v)) = v$. This system can be routinely solved using the matrices for the transformations $L_2$ and $L_1$.

To think about it in an intuitive way, we want a vector for which the rotation about $e_3$ moves $v$ to $L_2(v)$, and the rotation about $e_1$ moves $L_2v$ back to $v$.

As it turns out, the correct vector to choose is $$ v = (1,1,-1) $$ (or any non-zero multiple of this vector). Verify that $$ L_2(v) = (1,-1,-1)\\ L_1(L_2(v)) = L_1(1,-1,-1) = (1,1,-1) $$

$\endgroup$
6
  • $\begingroup$ but axis of rotation has also a direction (two possibilities) , how do you solve this problem from eigenvectors ? $\endgroup$
    – Widawensen
    Apr 25 '18 at 12:57
  • $\begingroup$ @Widawensen didn't consider that. Also, I now realize that I missed that $L_2$ is clockwise. $\endgroup$ Apr 25 '18 at 12:59
  • $\begingroup$ @Widawensen actually, I'm betting that the book posing this question doesn't care about the axis direction; note that it refers to a clockwise rotation about the $e_3$-axis rather than a counterclockwise rotation about the $-e_3$-axis, which would be the correct "direction" for the axis of that rotation. $\endgroup$ Apr 25 '18 at 13:05
  • $\begingroup$ If direction is not important then eigenvector is sufficient. However by changing the direction of the axis we are substancially changing the final result of rotation as convention for +/- angle is valid only with directed axis ( right hand rule) .. $\endgroup$
    – Widawensen
    Apr 25 '18 at 13:08
  • $\begingroup$ @Widawensen true, but note that they're not asking for the angle of the rotation $\endgroup$ Apr 25 '18 at 13:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.