So I am doing some exercises in Susan Lea's book Mathematics for Physicists, currently question 7 chapter 1:

  1. Find the matrix that represents the transformation obtained by (a) rotating about the x-axis by $45^\circ$ counterclockwise and then (b) rotating about the y'-axis by $30^\circ$ clockwise.

This is the relevant part of the question. The first part I get the correct rotation matrix. For the second part I get this rotation matrix:


which is equivalent to :


Whereas the book's solution is:

$\left[\begin{array}{2}\frac{\sqrt3}{2}&0&-\frac{1}{2}\\0&1&0\\ \frac{1}{2}&0&\frac{\sqrt3}{2}\end{array}\right]$

I believe there may be something wrong in my understanding. I am using the formula for a rotation matrix:

$\mathbf A _{ij} = cos(\theta_{ij})$

where $\theta_{ij}$ is the angle between the ith new axis and jth axis in the original system.

Can someone explain where I went wrong if I did.


closed as off-topic by AccidentalFourierTransform, stafusa, Kyle Kanos, Jon Custer, Cosmas Zachos Aug 15 '18 at 15:13

This question appears to be off-topic. The users who voted to close gave this specific reason:

  • "Homework-like questions should ask about a specific physics concept and show some effort to work through the problem. We want our questions to be useful to the broader community, and to future users. See our meta site for more guidance on how to edit your question to make it better" – AccidentalFourierTransform, stafusa, Kyle Kanos, Jon Custer, Cosmas Zachos
If this question can be reworded to fit the rules in the help center, please edit the question.

  • $\begingroup$ I've never seen that formula for the elements of the rotation matrix before, is it from that book or is it found somewhere else?? $\endgroup$ – N. Steinle Aug 12 '18 at 22:25
  • $\begingroup$ It is from the book listed as well as in Thornton and Marion's Classical Dynamics of particles and systems 4th edition. $\endgroup$ – QuantumPanda Aug 12 '18 at 22:48
  • 2
    $\begingroup$ Hi. You are rotating clockwise on y-axis so you rotate it as though the direction of the y-axis is the direction of the front face of clock. So look at initial angle of axis $x$ to final axis $z$ in element $A_{13}$. Its $90^0$ (initial xz angle) + the rotation $30^0$ clockwise = $120^0$. Same as element $A_{31}$, initial z to final x, now $60^0(90^0-30^0)$. $\endgroup$ – philip_0008 Aug 12 '18 at 22:53
  • $\begingroup$ @N.Steinle - It's correct. It's why an alternative name for a rotation/transformation matrix is the direction cosine matrix, and the elements of such a matrix are called direction cosines. One has to take great care, however, in considering which is which. $\endgroup$ – David Hammen Aug 13 '18 at 4:32
  • $\begingroup$ @QuantumPanda - You cannot and must not use the concepts from one textbook in answering questions from another. There are several arbitrary choices when it comes to rotation / transformation matrices, and between any two textbooks, the odds are strong that Murphy's Law applies. $\endgroup$ – David Hammen Aug 13 '18 at 5:56

Your matrix is the right one. The matrix given by Susan Lea is incorrect.

The matrix for the transformation of coordinates from an initial cartesian right-handed system of coordinates $\;Ox_1x_2x_3\;$ to another cartesian right-handed system of coordinates $\;Ox'_1x'_2x'_3\;$ rotated with respect to the first around a direction $\;\mathbf{n}=\left(n_1,n_2,n_3\right),\Vert\mathbf{n}\Vert=1, \;$ through an angle $\;\theta\;$ (positive if counterclockwise with respect to $\;\mathbf{n}$) is(1)

\begin{equation} \mathbb{A}\left(\mathbf{n}, \theta\right)= \begin{bmatrix} \cos\theta+(1-\cos\theta)n_1^2&(1-\cos\theta)n_1n_2+\sin\theta n_3&(1-\cos\theta)n_1n_3-\sin\theta n_2 \vphantom{\dfrac12}\\ (1-\cos\theta)n_2n_1-\sin\theta n_3&\cos\theta+(1-\cos\theta)n_2^2&(1-\cos\theta)n_2n_3+\sin\theta n_1\vphantom{\dfrac12}\\ (1-\cos\theta)n_3n_1+\sin\theta n_2&(1-\cos\theta)n_3n_2-\sin\theta n_1&\cos\theta+(1-\cos\theta)n_3^2\vphantom{\dfrac12} \end{bmatrix} \tag{01}\label{eq01} \end{equation}

So, the transformation matrices around axes $\;x_1,x_2,x_3\;$ through angles $\;\theta_1,\theta_2,\theta_3\;$ respectively are \begin{align} \mathbb{A}\left(x_1, \theta_1\right)& = \begin{bmatrix} \hphantom{-}1 & \hphantom{-}0 & 0 \vphantom{\dfrac12}\\ \hphantom{-}0 & \hphantom{-}\cos\theta_1 & \sin\theta_1\vphantom{\dfrac12}\\ \hphantom{-}0 & -\sin\theta_1 & \cos\theta_1\vphantom{\dfrac12} \end{bmatrix} \tag{02.1}\label{eq02.1}\\ \mathbb{A}\left(x_2, \theta_2\right)& = \begin{bmatrix} \cos\theta_2 & \hphantom{-}0 & -\sin\theta_2\vphantom{\dfrac12}\\ 0 & \hphantom{-}1 & \hphantom{-}0 \vphantom{\dfrac12}\\ \sin\theta_2 & \hphantom{-}0 & \hphantom{-}\cos\theta_2\vphantom{\dfrac12} \end{bmatrix} \tag{02.2}\label{eq02.2}\\ \mathbb{A}\left(x_3, \theta_3\right)& = \begin{bmatrix} \hphantom{-}\cos\theta_3 & \sin\theta_3 & 0 \hphantom{-} \vphantom{\dfrac12}\\ -\sin\theta_3 & \cos\theta_3 & 0 \hphantom{-}\vphantom{\dfrac12}\\ \hphantom{-}0 & \hphantom{-}0 & 1\hphantom{-} \vphantom{\dfrac12}\\ \end{bmatrix} \tag{02.3}\label{eq02.3} \end{align}

After a careful look in above equations we note that in the cases of rotation around $\;x_1,x_3\;$ [equations \eqref{eq02.1},\eqref{eq02.3}] the term $\;\boldsymbol{+}\sin\theta_\jmath\;$ is up-right while the term $\;\boldsymbol{-}\sin\theta_\jmath\;$ is bottom-left. But in the case of $\;x_2$, equation \eqref{eq02.2}, we have the inverse : the term $\;\boldsymbol{+}\sin\theta_3\;$ is bottom-left while the term $\;\boldsymbol{-}\sin\theta_3\;$ is up-right. In this difference lies the error in Susan Lea, see Figure below (page extracted from $^{\prime}$Student Solutions Manual for Mathematics for Physicists$^{\prime}$ by Susan Lea)

enter image description here

The corrected matrix $\;\mathbb{A}_2\;$ must be \begin{equation} \mathbb{A}_2= \begin{pmatrix} \cos\left(-30^\circ\right) & \hphantom{-}0 & -\sin\left(-30^\circ\right)\vphantom{\dfrac12}\\ 0 & \hphantom{-}1 & \hphantom{-}0 \vphantom{\dfrac12}\\ \sin\left(-30^\circ\right) & \hphantom{-}0 & \hphantom{-}\cos\left(-30^\circ\right)\vphantom{\dfrac12} \end{pmatrix} = \begin{pmatrix} \hphantom{-}\frac{\sqrt{3}}{2} & \hphantom{-}0 & \hphantom{-}\frac{1}{2}\vphantom{\dfrac12}\\ \hphantom{-}0 & \hphantom{-}1 & \hphantom{-}0 \vphantom{\dfrac12}\\ -\frac{1}{2} & \hphantom{-}0 & \hphantom{-}\frac{\sqrt{3}}{2}\vphantom{\dfrac12} \end{pmatrix} \tag{03}\label{eq03} \end{equation}

(1) For the matrix of equation \eqref{eq01} see equation (08) in my answer there : Rotation of a vector. But since this matrix of equation (08) therein concerns the rotation of a vector (active view) we replace $\;\theta\;$ by $\;\boldsymbol{-}\theta\;$ to get the matrix \eqref{eq01} herein in order to have the transformation of coordinates (passive view).

  • $\begingroup$ This does not provide an answer to the question. To critique or request clarification from an author, leave a comment below their post. - From Review $\endgroup$ – honeste_vivere Aug 13 '18 at 18:39
  • $\begingroup$ @ honeste_vivere : Iam preparing a full answer. Be patient. $\endgroup$ – Frobenius Aug 13 '18 at 18:49
  • $\begingroup$ Isn't the difference in whether or not the rotation is active or passive? I wouldn't say it is incorrect, since the problem does not specify. $\endgroup$ – Aaron Stevens Aug 13 '18 at 19:45
  • $\begingroup$ @Aaron Stevens : It concerns the rotation of coordinates axes (passive) not the rotation of points in a fixed frame of axes. $\endgroup$ – Frobenius Aug 13 '18 at 19:50
  • $\begingroup$ You are right I misread the question. I need to stop trying to multitask! $\endgroup$ – Aaron Stevens Aug 13 '18 at 20:09

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