Say we have a stick on length $L$ that travels toward a point $P$ with linear velocity $v$ in the direction of $P$ and rotational velocity $\omega$. The point and the stick are initially $l$ apart. Will a point $Q$ on the stick, a distance $r$ from its center, hit $P$?

Let the starting position of the stick be such that it points along the $x$-axis and that the center of the stick is at the origin. Looking from the point of view of the stick, the point $P$ will have a velocity $-v$ towards the $x$-axis (and the stick). The initial position of $P$ is $(d,l)$.

The equations of motion are thus \begin{align} \overline{P}&=\begin{pmatrix} d \\ l-vt \end{pmatrix} \\ \overline{Q}&=r\begin{pmatrix}\cos(\omega t) \\ \sin(\omega t) \end{pmatrix} \end{align}

and the question becomes "when is there a $t$ such that the two are equal?"

Introducing $\gamma\equiv\arccos\frac{d}{r}$ (which clearly shows the constraint $r\geq \vert d\vert$), we have

$$\frac{l-\gamma \frac{v}{\omega}}{r}=\sin \gamma$$

However, this looks transcendental.

Is there really no way of analytically figuring out whether or not $Q$ hits $P$? Can we determine whether or not the stick hits at all (for any $Q$)?

Edit: Just an idea: Perhaps one can use the Intermediate value theorem to show that there must exist a solution given some initial conditions?

Edit: Perhaps the question should be "For which $r$s (or some other, single parameter) is $\frac{l-\gamma \frac{v}{\omega}}{r}=\sin \gamma$ true (given $d,\ell,v,\omega$)? Can we say anything sensible about this?

  • $\begingroup$ There is certainly a complicated region of space that $P$ can be in where the stick entirely misses it: i.stack.imgur.com/B5owf.png $\endgroup$ – Rahul Sep 11 '16 at 21:53
  • $\begingroup$ @Rahul That is a beautiful illustration! Where did you find it/how did you make it? And does the space that the stick traces out have a name, i.e. is it some kind of well-known function? $\endgroup$ – Bobson Dugnutt Sep 11 '16 at 22:12
  • $\begingroup$ It's just a plot of $Q-P$ for fixed $d,l,v,\omega$ and variable $t,r$. I made it using the ParametricPlot command in Mathematica. $\endgroup$ – Rahul Sep 11 '16 at 22:18

$$ \frac{\ell - \gamma \frac{v}{\omega}}{r} = \sin\gamma. \tag{*} $$ Perhaps the problem is: Given $\ell,r,v,\omega$, solve for $\gamma$.

This is a transcendental equation. In fact, it is "Kepler's equation", and it is known that the solution (although unique) cannot be written in terms of "elementary functions".

LINK gives Kepler's equation as $$ M=E-e\sin E \tag{K}$$ where we should solve for $E$, given $M$ and $e$. We may substitute $$ E=\gamma \\ e=\frac{-r\omega}{v} \\ M=\frac{\ell \omega}{v} $$ to convert (K) into (*).

That link also discusses algorithms for numerical and series solution of (K).

  • $\begingroup$ Thank you for your answer. Note that $\gamma(r)$. Also, please have a look at my latest edit - I'm sorry my question was not clear. $\endgroup$ – Bobson Dugnutt Sep 11 '16 at 22:12

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.