Problems with two interacting objects can be simplified
by combining them into a pseudo-object with reduced massμ,
whose position equals the relative position of the objects.
For bodies 1 and 2 with respective masses m1 and m2:
μ≡m1+m2m1m2
If x1 and x2 are the objects’ respective positions,
then we define
the relative position xr,
the relative velocity vr,
and the relative acceleration ar:
Meanwhile, Newton’s third law states that
if object 1 experiences a force F1=m1a1 caused by object 2,
then object 2 experiences an opposite and equal force F2=−F1.
In fact, our earlier relation between a1 and a1
boils down to Newton’s third law:
F2=m2a2=−m1a1=−F1⟹a2=−m2m1a1
With all that in mind, let us take a closer look at the relative acceleration ar:
Where μ is the reduced mass, as defined above.
In other words, the relative acceleration ar
is just a1=F1/m1 multiplied by m1/μ.
This can be regarded as focusing on the dynamics of body 1,
while correcting for the effects of body 2.
This also suggests the following way
to recover the original positions x1 and x2
from xr, which you can easily verify for yourself:
Then, assuming that the system’s potential energy V
only depends on the distance between the two objects,
i.e. V=V(∣x1−x2∣)=V(∣xr∣),
we just showed that we can rewrite both T and V
to contain only μ and relative quantities.
This is relevant for both Lagrangian mechanics
and Hamiltonian mechanics,
where L=T−V and H=T+V respectively.