Uploaded by the

keplerproofs

advertisement
A1 Dynamical Astronomy
Proof of Kepler’s laws from Newtonian dynamics
It would be a pity to have a course on dynamical astronomy and not at least see a proof of Kepler’s laws
from Newton’s laws of motion and gravitation. Importantly, these proofs are not examinable! They
are presented here purely to satisfy curiosity and for
your entertainment.
Similarly, we can write the angular momentum of π‘š
as
(3)
𝐿 = π‘šπ‘Ÿ2 πœƒ,Μ‡
as π‘Ÿπœƒ Μ‡ is the component of 𝑣 perpendicular to π‘Ÿ. Now
make the substitution 𝜌 = 1/π‘Ÿ, so that πœƒ Μ‡ = 𝐿𝜌2 /π‘š.
We can evaluate πœƒ as
𝐿 2
𝜌 d𝑑
π‘š
𝐿 d𝑑
= ∫ 𝜌2
d𝜌
π‘š d𝜌
1 d𝜌
but π‘Ÿ Μ‡ = − 2
𝜌 d𝑑
𝐿
so πœƒ = − ∫
d𝜌.
π‘šπ‘Ÿ Μ‡
πœƒ=∫
KEPLER’S FIRST LAW
Ki: A planet orbits the Sun in an ellipse, with the Sun
at one focus of the ellipse. Take a mass π‘š in a general
π‘ŸΜ‡
π‘š
π‘Ÿπœƒ Μ‡
(4)
(5)
(6)
(7)
Rearranging Eqn. (2) we can see that π‘Ÿ Μ‡ is
π‘Ÿ
𝐿2
2𝐸
+ 2πΊπ‘€πœŒ − 2 𝜌2 .
(8)
π‘š
π‘š
Now a further manipulation. We make the substitutions
πœƒ
π‘Ÿ2Μ‡ =
𝑀
π‘Ÿ0
𝐿2
(9)
πΊπ‘€π‘š2
2πΈπ‘Ÿ0
(10)
𝑒2 = 1 +
.
πΊπ‘€π‘š
Figure 1: The geometry used in the proof.
Clearly, both π‘Ÿ0 and 𝑒 are constants. We choose them
in this way so that our answer is immediately recogelliptical orbit around a much more massive body 𝑀.
nisable as an ellipse—it is not an obvious substitution
When the separation of the masses is π‘Ÿ the total enat this stage! After a little manipulation, Eqn. (8) can
ergy of the orbit is
be written as
π‘Ÿ0 =
1
πΊπ‘€π‘š
𝐸 = π‘šπ‘£2 −
,
2
π‘Ÿ
2 1/2
(1)
𝐿 𝑒2
1
π‘Ÿ Μ‡ = [ 2 − (𝜌 − ) ]
π‘š π‘Ÿ0
π‘Ÿ0
where 𝑣 is the speed of the orbiting mass and 𝐺 the
constant of gravitation. The velocity of π‘š has two
components: a radial component equal to dπ‘Ÿ/d𝑑 (written π‘Ÿ)Μ‡ and a component perpendicular to π‘Ÿ which is
the ‘circular’ component of the velocity, equal to π‘Ÿπœ”
where πœ” (≡ πœƒ)Μ‡ is the instantaneous angular velocity
of the body, with πœƒ as shown in Fig. 1. Because these
components are orthogonal, the square of the total
velocity equals the sum of the squares of these components. We can now write this energy equation in
polar coordinates:
Substituting this into Eqn. (7) we get
1
πΊπ‘€π‘š
𝐸 = π‘š(π‘Ÿ2Μ‡ + π‘Ÿ2 πœƒ2Μ‡ ) −
.
2
π‘Ÿ
πœƒ = −∫
.
1
√(𝑒/π‘Ÿ0 )2 − (𝜌 − 1/π‘Ÿ0 )2
𝜌 − 1/π‘Ÿ0
= cos−1 (
).
𝑒/π‘Ÿ0
(11)
d𝜌
(12)
(13)
This can be rearranged to give
π‘Ÿ = π‘Ÿ0 /(1 + 𝑒 cos πœƒ),
(14)
which is the equation of an ellipse in polar coordinates, with the origin at a focus. We can now identify
π‘Ÿ0 as the semi-latus rectum of the ellipse and 𝑒 as its
(2) eccentricity.
A1 Dynamical Astronomy
KEPLER’S SECOND LAW
KEPLER’S THIRD LAW
Let’s now prove Kii: The line joining a planet to
the Sun sweeps out equal areas in equal intervals of
time. The proof of Kii highlights the generality of the
‘sweeping out area’ rule for motion under any central
force.
Now for Kiii: The square of the orbital period of a
planet is proportional to the cube of the semi-major
axis of its orbit. We’ll use some of the results from
the lectures for this one.
d𝒓
𝒓
𝐴tot = ππ‘Žπ‘ = ππ‘Ž2 √1 − 𝑒2 ,
𝛼
d𝐴
Figure 2: The area, d𝐴, swept out in a time d𝑑 by 𝒓.
In a time d𝑑 the planet will move by a small
amount d𝒓. The small triangle this vector makes to
the Sun (Fig. 2) has an area
1
d𝐴 = π‘Ÿdπ‘Ÿ sin 𝛼,
2
(15)
where 𝛼 is the angle between 𝒓 and d𝒓 (remember
the area of a triangle is ½ π‘Žπ‘ sin 𝐢). This can be usefully written as a (pseudo)vector perpendicular to the
plane of the triangle with magnitude d𝐴 using the
vector cross-product:
1
d𝑨 = 𝒓 × d𝒓.
2
The total area of an ellipse of semimajor and
semiminor axes π‘Ž and 𝑏 is
(19)
where 𝑒 is again the eccentricity (we derived 𝑏 =
π‘Ž√1 − 𝑒2 in the lectures using the ‘string’ definition
of an ellipse). From Eqn. (17), the rate of sweeping
out area is
1
𝑳
1
𝑨̇ = 𝒓 × π’“ Μ‡ = 𝒓 × π’— =
,
(20)
2
2
2π‘š
where 𝑳 is the planet’s orbital angular momentum
around the Sun and π‘š is its mass. The orbital period
𝑇 is simply the time taken to sweep out an area 𝐴tot ,
i.e.,
π‘š
ππ‘Žπ‘
= 2ππ‘Ž2 √1 − 𝑒2 ,
(21)
𝑇=
𝐿
𝐿/(2π‘š)
so
π‘š2
𝑇 2 = 2 4π2 π‘Ž4 (1 − 𝑒2 ).
(22)
𝐿
𝑑
π‘Ÿ0
2π‘Žπ‘’
(16)
The rate of sweeping out area due to movement is
Figure 3: Relating π‘Ÿ0 to π‘Ž and 𝑒.
therefore
d𝑨 1
We are nearly there, but we need to address the 𝐿
𝑨̇ =
= 𝒓 × π’“.Μ‡
(17)
d𝑑
2
and 𝑒 terms in this expression. Using Fig. 3, and that
Kepler’s second law states that this is a constant for 𝑑 + π‘Ÿ0 = 2π‘Ž for an ellipse, we have (by Pythagoras)
the orbital motion, so π‘¨Μˆ should be zero if Kii holds.
(2π‘Ž − π‘Ÿ0 )2 = 4π‘Ž2 𝑒2 + π‘Ÿ02
(23)
Differentiating with respect to time again gives
2
π‘Ÿ0 = π‘Ž(1 − 𝑒 ).
(24)
1
π‘¨Μˆ = (𝒓 Μ‡ × π’“ Μ‡ + 𝒓 × π’“)̈ .
(18) Inserting π‘Ÿ0 from Eqn. (9) we get
2
π‘š2
1
The first term on the right-hand side certainly equals
=
.
(25)
2
𝐿
πΊπ‘€π‘Ž(1 − 𝑒2 )
zero, as it is the cross-product of a vector with itself,
but the second term is not zero for general motion. We can therefore write Eqn. (22) as
However, 𝒓 ̈ is just the acceleration of the planet, and
4π2 3
π‘Ž ,
(26)
𝑇2 =
by Newton’s second law that is in the direction of the
𝐺𝑀
applied (gravitational) force, so is also directed along which is Kiii. As a bonus we get the constant of pro𝒓. Therefore the second term must also be zero in portionality.
this case. We can therefore say that π‘¨Μˆ = 0 and so
𝑨̇ is a constant. It’s clear this would be true for any
GW
‘central force’, where the force is directed along the
2020
v.557
line connecting the centres of mass.
Download