Dynamics I

12-Feb-2014
E. Schrama e.j.o.schrama@tudelft.nl

Overview
Part 1
! The two-body problem:
! Potential from a point mass

! The three-body problem:

! Uniform rotating system with two large planets

! Hill equations and Hill sphere ! Planetary perturbations and resonances

Part 2
! Tidal forcing and dissipation ! Tides in the solar system ! Roche limit and tidal energy dissipation ! Dissipative forces and the orbits of small particles

The two-body problem, sec 2.1

! A large planet and a small satellite
! Potential energy: "r " ! Total energy: 2a ! Angular momentum: !

a(1 " e2 )

! Keplers !laws on orbit motions

! Elliptical orbits ! within an orbital plane ! Equal area law ! Scale vs orbital period law

! Equations of Motion

Keplers first law

a(1 # e ) r(" ) = 1 + e cos(" )
There are 4 cases:
2

Figure center of ellipse (irrelevant for mechanics)

"

! e=0, circle ! 0<e<1, ellipse ! e=1, parabola ! e>1, hyperbola

Conic sections

focal point ellipse

In-plane Kepler parameters

r = a (1 + e)

r = a (1 ! e)

ae
Apo-apsis Peri-apsis

Keplers second law

A B

D O C

ABO ! CDO

Keplers Third law

n a = G( M + m)

2 3

2! T= n

The variable n represents the mean motion in radians per second, a is the semi-major axis, G is the gravitational constant, M is the mass of the Sun, m is the mass of the satellite (m << M) and T is the orbital period of the satellite In astronomy this law provides the scale of the Solar System, we can observe rather precisely the orbital periods of planets, and from this information we can infer the distances in the solar system. Everything is then normalized to the Earths orbital radius, which is said to be 1 astronomical unit (1 AU)

The three body problem, sec 2.2

! Configuration consists of three arbitrary masses that attract one another ! Take Newtons gravity Law and add up all the forces (and convert to accelerations) ! Define the barycenter of the system ! Only numerical solutions are tractable ! More common in astronomy is the restricted three body problem

Restricted three body problem

! Configuration consists of two large masses (mp and mq) that are about of the same size. ! In addition there is a small particle ! The barycenter is between mp and mq ! The system is rotating at a uniform rate ! Equations of motion include frame terms as a result of this rotation

Jacobi Integral
! The energy integral in a rotation frame is:

2mp 2 mq 2 CJ = x + y + + !v r ! rp r ! rq
2 2

! Where P and Q are planets, x, y are barycentric coordinates, all rs are vectors and v is a velocity in a plane that rotates about z. ! Lets try to understand this integral first

barycenter

Planet p Lagrange point Planet q

See also: http://janus.astro.umd.edu/javadir/orbits/ssv.html

Balance rotation and gravity

For mp and mq we have:

Uniform rotation
x2 !2 !1

!
x1

x = R3 (" )!

Equations of motion after rotation

After straightforward differentiation we get

If we ignore the Coriolis terms (where?), then we obtain the gradient of CJ

So that we can plot the length of the acceleration vector on the left hand side to demonstrate the existence of the Lagrangian points L1 till L5 !

mp = 10 mq = 1

Horseshoe and Tadpole orbits

Epimetheus Janus Saturn

Figure 2.5 in the book, all represented in co-rotating frame.

Exercises
! Find the exact locations of L1 to L5 (notes) ! Where could you put a satellite to warn us before the arrival of a solar storm? ! Where do we find the Trojan asteroids, which planets cause them to be there?

Intermezzo : some theory

! Where are the Lagrange points in the 3 body problem? ! What is the link to the Hill sphere radius? ! Discuss the Hill equations ! This part is not in the PS book, but there are notes on backboard that help you out. ! These notes on blackboard are part of this course.

Hill equation set-up

# " z y v u ! w

Some intermediate steps

For the equations of motion we always need the acceleration and the gradient of the potential (V)

(a.k.a. unfinished Hill equations)

Potential function V=U+T

! The U part belongs to a point mass (U=/r) ! T is called the disturbing potential ! We approximate V at the position of the satellite, thus (!,",") = (r+u,v,w) :

This last relation is now substituted in the unfinished version of the Hill equations

Hill equations
! For the rotating coordinate system one can show that:

! Here u, v and w are local coordinates for a satellite relative to a circular reference orbit (thus, u: radial, v: flight direction and w: normal to plane, and T is a perturbing potential whose partial derivates are required in the local co-rotating frame)

Solutions of the Hill Equations

! Homogeneous:

! = Fu # u F = Q"Q T # u (t ) = Qe
! Particular:
T ! u + F u = b (t ) with F = Q!Q " T ! T ! + !z = c (t ) Q (u + F u ) = Q b ( x) " z
" .( t !t 0 )

Q u (t0 )

Solutions of the Hill Equations (2):

For each ith equation in the reduced particular system you should apply a Laplace transform:

" i ( t ) = di e
t

# $i t

+e

# $i t

Ci ( t )

Ci ( t ) =

t0

%e

$i x

c i ( x ) dx

ui ( t ) = Q " i ( t )

Characteristic solution homogenous case

This example works well for rendezvous and docking type op situations

STS133 and ISS on 26-feb-2011

Characteristics particular solution

Corollary
! We apply an acceleration function at frequency $ ! The orbit is perturbed at frequency $ ! Resonances appear when $ is near n ! It works in the same way as mass spring systems do ! Read section 2.3

Example problem 1
! Derive the analytical solution of the homogeneous solution of the Hill equations, and substitute initial conditions ! Explain how you can solve a rendez-vous between the Space shuttle and a Space Station with the obtained homogeneous solution of the Hill equations.

Hill Sphere
! The approximate limit to a secondarys (Moons or planets) gravitational dominance is given by the extent of its Hill sphere:
1/ 3 # & mq RH " a% % 3( m + m ) ( ( $ p q '

! !

A test particle located at the boundary of the Hill sphere is subject to a gravitational force from the planet comparable to the tidal difference between the force of the Sun on the planet and that on the test body The Hill sphere stretches out to Lagrangian point L1 All planetocentric orbits within the Hill sphere are stable over long periods of time, this is where we find all natural satellites.

Section 2.2.3 See notes on BB for L2 to L5

Why is Rh like shown?

! For this we turn back to the Hill equations ! For L1 we have that only the u equation counts, v and w are zero, and there are no perturbations (T=0). The u equation becomes:

!! = u
3 h

G.mq R
2 h

= 3n u = 3
3

G.( m p + mq ) a
3

Rh
1/ 3

& # mq R = a ' Rh = a.$ ! 3(m p + mq ) $ % 3(m p + mq ) ! "

mq

Example problem 2
! Show that our Moon is within the Hill sphere of the Earth ! Currently the Moon disappears at a rate of 3.8 cm per year due to tidal energy dissipation.
! How long would it take before we lose our Moon because it leaves our Hill sphere? ! You should find that this happens after % Gy which is roughly % times the age of our solar system

Example problem 3
! If the Hill radius is set to be at the outer Oort cloud radius (105 AU) and if we are at 27000 light-years from the center of the galaxy, then what is the weight of the galaxy expressed in solar masses? ! You should find roughly %. solar masses while 200-600 billion solar masses are optically estimated.

Dynamics II
23-Feb-2014

E. Schrama e.j.o.schrama@tudelft.nl

Planetary Perturbations and Resonances

Within the solar system all planets and Moons attract one another causing their Kepler elements to change over time. Susceptible elements are for instance the semi-major axis, the inclination and the eccentricity of asteroids and comets. Planetary orbit perturbations may be approximated by a second order system that is perturbed by an oscillation (see Hill equations):

+ n w = A cos" t + B sin " t w

The solution of this system is that of a harmonic oscillator:

A cos "t + B sin "t w ( t ) = C0 cos nt + C1 sin nt + n2 # " 2

This solution is resonant when n approaches $ (what does it mean) Section 2.3.1

Examples of resonances in the solar system

! Epimetheus-Janus motion (see Lagrange point discussion). ! Discovery of Neptune via Saturns orbit by Urbain Le Verrier (albeit that it is surrounded by controversy) ! Resonances may force material into highly eccentric orbits, this results in collisions which clears the Kirkwood gaps in the asteroid belt.

Compare to fig 12.8 in Planetary Sciences book

(JPL)

Compare to fig 12.8 in Planetary Sciences book

(JPL)

Long term stability of planetary orbits

! Why is the solar system currently as is? The system is 4.5 billion years old (comes from meteorite observations) ! Are there periods of chaotic behavior in the solar system, how did the system emerge out of a proto-disc? ! Analytically you get resonances, but they are not real, because you are approximating the real dynamics. ! Numerical integration shows that the system remains stable for millions of years, accretion is important to explain the formation. ! The Earths orbital elements are harmonic expressions, and there are 40k and 100k year variations in the orbital elements. The variations are related to the Milankovitch cycles, it allows you to compute insolation functions.

Mean annual insolation

Mean daily insolation at 65N

(Laskar, 1993)

Orbits about an oblate planet

! Keplers theory is good approximation of reality, and it describes the motion in gravity fields near large bodies like Earth or Sun pretty well. ! The potential of the field is what we called U for the Kepler problem, but, there are higher order moments which we ignored ! The complete gravitational potential V is formally expressed as V=U+T, where T is a disturbing potential. ! Gravitational flattening (a part of T) results in a special situation which does significantly affect orbits, and the result can be explained analytically ! In the following many mathematical details are skipped, we only show the solution! ! Sections 2.5 and 2.6

Perturbed equations of motion

! ! i X = *+V + , F
i

GM GM V =* * r ae

) ae & ' $ CnmaYnma (/ , - ) , n ,m,a ( r % a=0 a =1

n +1

#cos m- : Ynma (/ , - ) = Pnm (cos . )" ! sin m- :

Typical questions are: 1) what coordinate system do we use? 2) explain all symbols 3) what are the ranges of the indices n,m and a 4) calculate the gradient of V

Solution equations of motion

Analytic
! Lagrange planetary equations ! Gravity potential in Keplerian elements ! Isolate the first-order solution ! Approximate higher-order perturbations

Numeric
! Conversion to system of first-order ODE ! Integration of system of equations

Linear perturbations by C20

d" 3na = C20 2 cos I 2 2 dt 2a (1 ! e ) 2 d# 3nae 2 = C20 2 (1 ! 5 cos I ) 2 2 dt 4a (1 ! e ) 2 dM 3nae 2 = n ! C20 2 (3 cos I ! 1) 2 3/ 2 dt 4a (1 ! e )
Notes: ! C20 not normalised, n: mean motion ! Similar expressions for higher-order moments
2 e

J2/Jodd resonance

J2/Jodd resonance

Example for Cryosat-2

C/k

Tides, section 2.7

! Newton once said:
! A tidal force is caused by the gravitational attraction of a remote body on your planet. ! The DIFFERENCE between the local force and the force acting on the center of gravity of your planet is relevant for tides

! Our tides are primarily caused by Sun and Moon ! In reality everything else in the solar system matters in the exact formulation of tidal accelerations.

Tide generating force

There are tidal potential expressions where the gradient results in the tidal acceleration vector discussed here.

Tidal force or potential

! Acceleration of force vectors is easy to understand, but difficult to implement. ! We define a potential function U ! U is scalar, you can add different Us ! Gradient of U is an acceleration vector ! Gradient of tidal potential results in the tide generating force by unit of mass ! The proof goes via work integral method (separate notes on blackboard)

Tidal potential
n " % m Re a U = $ ' Pn (cos * ) ) Rem n=2 # Rem & (

Re

!
Rem

Re

Ua =

s =0

# ("a , n )ds
,cos +/ . 1 n = . sin + 1 . - 0 1 0

%2 0 0 (,"x1 = s.cos +/ *. m ' 1 "a = 3 '0 $1 0 *. "x2 = s.sin + 1 and Rem ' *. 0 0 $ 1 0 & )- "x3 = 0 1
Re

2 R 2 m Ua = # s 3 3cos + $ 1}ds = m 3 e P2 (cos + ) { Rem s =0 Rem Include third and higher order derivatives : n % ( m Re a U = ' * Pn (cos + ) 3 Rem n=2 & Rem ) 2

Laplace equation and Legendre functions

! If you start with &U=0 and if you apply the method of separation of variables then
! U(r,',() = R(r) G(',() ! R(r) = c1rn + c2r-(n+1) ! G(',() = [Anmcos(m') + Bnmsin(m')] Pnm(cos()

! Appendix A tides notes is entirely on properties of Legendre functions

! Zonal and tesseral spherical harmonics ! orthogonality ! Additional theorem ! convolution integral

Tidal Energy Dissipation

! Tides are generated by the Sun and the Moon ! Continuous forcing shouldnt result in continuous amplification, somehow a wave has to lose its energy ! Conversion of mechanical energy into heat or another form of energy is called dissipation ! Bottom friction is such a process, and for long it was assumed this is the only way it works ! Global energy dissipation studies: purpose, provide one number like 2.5 TW for M2 for the entire planet ! Local energy dissipation studies: purpose, provide one number for e.g. a coastal sea. ! Local dissipation may be related to internal tides

Tidal torques (1)

3.82 cm/yr M2 : 2.50 +/- 0.05 TW

Study section 2.7

Tidal torques (2)

! Due to the fact that mass motions on the planet are slower than planetary rotation, we get a phase lag between the bulge and the planet-moon line. ! Tidal bulges (in the ocean and interior) cause a torque on the planet, resulting in a slowdown in the rotation ! The moon sees a part of this effect, but it sees and acceleration in the flight direction, therefore our Moon recedes at a rate of 3.82 cm/yr ! All this in ONLY true for the configuration as shown for the Earth Moon system, it does not work for Mars and Phobos, something to figure out yourself. Phobos is inbound, it a doomed moon.

Tidal energy dissipation

! The existence of the amplitude of the tidal bulge (D) and its phase lag ()) is sufficient to derive the rate of energy loss. ! Energy loss (or more precise, energy conversion) is what we call dissipation ! How: gravitational forces set mass in motion (tides), mechanical is then converted into heat or chemical reactions or results in mixing ! Dissipation can be estimated directly from D and ), the latter two are normally observed

Global energy dissipation

+ * ' $ cos & + 2m < D >= 4 "R 2 # (1 + k 2 $ h2 ) % A% D 2m 2 m ) + , (+ sin& 2 m + Dnm cos& 2 m =

(anm ! dnm ) 1 Dnm sin& 2 = m 2 (c nm bnm )

1 2

Consequence Dissipation (Earth)

! Since that there is radar altimetry to our oceans we know precisely dissipation values at different tidal frequencies (M2 and S2 hereafter as examples) ! M2: oceanic 2.42, astronomic 2.51 TW, the difference is dissipated in the solid Earth tide (Ray, Eanes and Chao, 1996). Independent body tide dissipation measurements by gravimeters are not convincing at the moment (only a 0.1 of a degree lag is expected) ! S2: oceanic 0.40, geodetic 0.20 TW, the difference is mostly dissipated in the atmosphere (Platzman,1984)

Tidal approximation for a thin water layer on earth

M2 ocean tide

Tidal energy dissipation map

Earth Moon system

! Our Moon lost most of its eigenrotation since it was formed 4.5 billion years ago, it is currently phase locked to the Earth ! On Earth eigenrotation still exists but we experienced a slowdown (daylength was shorter in the past, and it long get longer in the future) ! On a geologic time-scale, ocean tides decreased in amplitude.

Jupiter system
! The moon Io experiences so much tidal heating that it is volcanic (BV lecture) ! Europa and Ganymede have frozen surfaces, with cracks and ejected material on the surface emerging from those cracks ! We know little about the subsurface of Europa, it may be slush or water ! What do you need to find out: e.g. altimetry on the new to fly Juice mission

Europas surface

Roche limit
! Near a large planet, tidal forces grow beyond the maximum stress limit of a body causing it to break up, this limit is called the Roches limit. ! The Roche limit is found by balancing the Moons self gravity at the surface against the tidal force of a planet

How large is the Roche limit

atides = abinding
rp d rs rs

2 p d
3

(2rs ) =

s
(2rs )
2

)
1/ 3

& *p # d ( 2.52 ' rp ' $ $* ! ! % s"

Some remarks
! Most Moons around planets are outside the Roche limit and within the Hill sphere ! The tensile strength of the Moon is not taken into account, but it does play a role ! There are Moons, like Amalthea around Jupiter and Pan around Saturn, that orbit within the Roche Limit.

Amalthea

Pan

Dissipative forces and the orbits of small satellites

! Radiation pressure (2.8.1)
! Sun light pushes particles (mostly micrometer sized dust) outwards

! Poynting-Roberton drag (2.8.2)

! Centimeter sized particles spiral inwards to the Sun

! Yarkovsky effect (2.8.3)

! Changes orbits of meter to km sized objects due to an uneven temperature distribution at the surface

! Corpusclar Drag (2.8.4)

! Like Poynting-Robertson, except that it is caused by solar wind particles

! Gas Drag (2.8.5)

! Gas (atmosphere or in interplanetary space) exerts a drag effect on a body proportional to the velocity squared of the body relative to the gas

! Orbits about a star that loses mass (section 2.9)

Radiation pressure

Frad

!A " 2 Q pr r 4 # cr

Q pr Frad &5 $= = 5.7 % 10 Fg 'R Fg, eff &(1 & $ )GmM = r2

Poynting Robertson Drag

c v

For particles that re-radiate pressure the net effect is a drag in the flight direction.

P0

(1 + v / c )P0

(1 ! v / c) P0

In Particles Frame

In Rest Frame of Sun

Yarkovski Effect

Gas drag

1 2 F = C d " Av 2

dE d # " & d # F . s & F = % (= % *v () dt dt $ 2 a ' dt $ M ' M

(use this for computing orbital decay)

Reference
! Planetary Sciences book
! Chapter 2, dynamics, 2.1 till 2.8 ! Chapter 13, planetary rings, 13.1

! PDF notes on blackboard

! Note on orbital mechanics ! Note on Hill equations ! Note on Roche Limit ! Note on Lagrange points