1
$\begingroup$

There is a well-known $N$-body problem in classical mechanics: Given an initial positions and velocities of $N$ particles in some space, describe their dynamics over some time interval.

I'm interested in some form of "inverse" problem: Let's assume that we know that there are $(N+1)$ particles in some space. We are given the trajectories of $N$ of these particles over some time interval. The problem is to restore the trajectory of the $(N+1)$-th particle over the same time interval.

The underlying force field is assumed to be known. For example, we can assume that each pair of particles is attracted according to the inverse square law.

What is a correct generally accepted name for this problem? Is this problem described in literature?

$\endgroup$
4
  • $\begingroup$ Are you assuming that the $(N+1)$th particle has infinitesimally small mass? $\endgroup$
    – Qmechanic
    Commented Nov 30, 2020 at 22:55
  • $\begingroup$ @Qmechanic♦ No, I don't make such assumption. Is it even meaningful to make such assumptions in newtonian mechanics? Would not it imply that the force acting on the $(N+1)$th body is zero? $\endgroup$ Commented Nov 30, 2020 at 23:09
  • $\begingroup$ Are you given initial conditions for the $N+1$st particle? If so then the first $N$ just define a time dependent potential $V(x,t)$ and you have a 1 body problem. $\endgroup$
    – jacob1729
    Commented Dec 1, 2020 at 0:09
  • $\begingroup$ @jacob1729 Let's assume that the initial conditions for the $(N+1)$th particle are unknown. $\endgroup$ Commented Dec 1, 2020 at 0:59

1 Answer 1

3
$\begingroup$

If the system is isolated, then the center of mass of this system moves at a constant (usually zero) velocity $\mathbf{v}_c$: $$ \sum_{i=1}^{N+1}m_i\mathbf{r}_i(t)=M \mathbf{r}_c(t)=M(\mathbf{r}_0+\mathbf{v}_c t) $$ If $\mathbf{r}_i(t)$ are known for all $i=1,\ldots,N$, then $\mathbf{r}_{N+1}(t)$ can be obtained from this equation: \begin{equation}\tag{1} \mathbf{r}_{N+1}(t)=\frac{1}{m_{N+1}}\left(M(\mathbf{r}_0+\mathbf{v}_c t)-\sum_{i=1}^{N}m_i\mathbf{r}_i(t)\right) \end{equation} This equation contains 2 unknown parameters: initial position of the center of mass $\mathbf{r}_0$ and its velocity $\mathbf{v}_c$. These parameters can be (presumably) obtained by requiring that the equations of motion hold (since the law of interaction is known).

UPDATE:

To obtain $\mathbf{r}_0$ and $\mathbf{v}_c$ from the equations of motion:

Assume that the potential energy is: $U=\sum_{i=1}^{N}\sum_{k=i+1}^{N+1}U(|\mathbf{r}_i-\mathbf{r}_k|)$. Then the equation of motion for each particle is: $$ m_i\mathbf{\ddot{r}}_i=-\sum_{k=1}^{N+1}U'(|\mathbf{r}_i-\mathbf{r}_k|)\frac{\mathbf{r}_i-\mathbf{r}_k}{|\mathbf{r}_i-\mathbf{r}_k|} $$ For the first particle: $$\tag{2} m_1\mathbf{\ddot{r}}_1=-\sum_{k=1}^{N}U'(|\mathbf{r}_1-\mathbf{r}_k|)\frac{\mathbf{r}_1-\mathbf{r}_k}{|\mathbf{r}_1-\mathbf{r}_k|}-U'(|\mathbf{r}_1-\mathbf{r}_{N+1}|)\frac{\mathbf{r}_1-\mathbf{r}_{N+1}}{|\mathbf{r}_{1}-\mathbf{r}_{N+1}|} $$ Substituting solution (1) into equation (2) and setting $t=0$ leads to an equation for $\mathbf{r}_0$. Of course, the equation can be nonlinear and it can have multiple solutions.

After $\mathbf{r}_0$ is found, $\mathbf{v}_c$ can be obtained from the same equation (2) for $t>0$.

$\endgroup$
4
  • $\begingroup$ Thank you for your answer! It would be great, if you could elaborate a little bit more on how $\mathbf{r}_0$ and $\mathbf{v}_c$ can be obtained using the equations of motion. $\endgroup$ Commented Dec 1, 2020 at 0:28
  • $\begingroup$ I have updated the answer $\endgroup$
    – atarasenko
    Commented Dec 1, 2020 at 0:59
  • $\begingroup$ Thank you so much. By the way, do I understand correctly that in general the $(N+M)$ problem (when trajectories for $N$ particles are known and for $M$ particles are unknown) cannot be solved using the same method? And presumably it doesn't have an analytical solution at all. $\endgroup$ Commented Dec 1, 2020 at 1:24
  • $\begingroup$ Yes, $N+M$ problem doesn't have a solution in general. For example, for $N=1$ and $M=2$; the first particle (with known trajectory) is aways at the point $\mathbf{r}_1(t)=0$, and two other particles (with unknown trajectories) are at the opposite points: $\mathbf{r}_2(t)=-\mathbf{r}_3(t)$. It is not possible to find $\mathbf{r}_2(t)$ in this problem because the forces acting on the first particle always compensate each other. It will stay at the origin for any $\mathbf{r}_2(t)$. $\endgroup$
    – atarasenko
    Commented Dec 1, 2020 at 1:39

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