Level 3 — ProductionOrbital Mechanics & Astrodynamics

Orbital Mechanics & Astrodynamics

45 minutes60 marksprintable — key stays hidden on paper

Level: 3 — Production (from-scratch derivations, code-from-memory, explain-out-loud) Time limit: 45 minutes Total marks: 60

Use ...... / ...... notation for all mathematics. Constants (unless given): μ=GM=3.986×105 km3/s2\mu_\oplus = GM_\oplus = 3.986\times10^{5}\ \text{km}^3/\text{s}^2, R=6378 kmR_\oplus = 6378\ \text{km}.


Question 1 — Orbit equation from scratch (12 marks)

Starting from the two-body equation of motion r¨=μr3r\ddot{\mathbf r} = -\dfrac{\mu}{r^3}\mathbf r:

(a) Show that specific angular momentum h=r×r˙\mathbf h = \mathbf r \times \dot{\mathbf r} is conserved. (3)

(b) By taking r¨×h\ddot{\mathbf r}\times\mathbf h and integrating, derive the orbit equation r=h2/μ1+ecosθ.r = \frac{h^2/\mu}{1 + e\cos\theta}. State clearly what the constant of integration (eccentricity vector) represents. (7)

(c) State the eccentricity ranges giving a circle, ellipse, parabola, and hyperbola. (2)


Question 2 — Vis-viva and Kepler III derivation (10 marks)

(a) Using conservation of energy ε=12v2μr\varepsilon = \tfrac12 v^2 - \tfrac{\mu}{r} and the fact that for an ellipse ε=μ2a\varepsilon = -\dfrac{\mu}{2a}, derive the vis-viva equation v2=μ(2r1a)v^2 = \mu\left(\dfrac{2}{r}-\dfrac{1}{a}\right). (4)

(b) Using Kepler's second law (areal velocity =h/2= h/2) and the ellipse area πab\pi a b, derive Kepler's third law T2=4π2μa3T^2 = \dfrac{4\pi^2}{\mu}a^3. (6)


Question 3 — Hohmann transfer, full numeric (12 marks)

A spacecraft is in a circular LEO of radius r1=6778 kmr_1 = 6778\ \text{km} and transfers via a Hohmann transfer to GEO at r2=42164 kmr_2 = 42164\ \text{km}.

(a) Compute the circular speeds v1v_1 and v2v_2. (2)

(b) Compute the transfer-ellipse semi-major axis and the perigee/apogee speeds on the transfer orbit. (4)

(c) Compute both burns Δv1\Delta v_1, Δv2\Delta v_2 and the total Δv\Delta v. (4)

(d) Compute the transfer time (half the transfer-orbit period). (2)


Question 4 — Kepler's equation solver, code from memory (10 marks)

(a) Derive Kepler's equation M=EesinEM = E - e\sin E starting from the definition of mean anomaly and eccentric anomaly (state the geometric/area argument). (4)

(b) Write pseudocode (or Python) for a Newton–Raphson solver that returns EE given MM and ee, including the iteration formula and a sensible initial guess and stopping criterion. (4)

(c) For M=0.8 radM = 0.8\ \text{rad}, e=0.3e = 0.3, perform one Newton iteration from E0=ME_0 = M and give the improved estimate. (2)


Question 5 — J2 nodal precession & Sun-synchronous orbit (10 marks)

(a) State the J2J_2 nodal regression rate Ω˙=32J2nR2a2(1e2)2cosi,\dot\Omega = -\frac{3}{2}\frac{J_2 n R_\oplus^2}{a^2(1-e^2)^2}\cos i, and explain physically why oblateness (J2>0J_2>0) causes the node to move, and why Ω˙=0\dot\Omega=0 at i=90°i=90°. (4)

(b) Explain what condition a Sun-synchronous orbit must satisfy and the sign/direction of inclination required to achieve it. (3)

(c) A satellite at a=7078 kma = 7078\ \text{km}, e=0e = 0 requires Ω˙=+1.991×107 rad/s\dot\Omega = +1.991\times10^{-7}\ \text{rad/s} (one revolution per year). Given J2=1.0826×103J_2 = 1.0826\times10^{-3}, find the required inclination ii. (3)


Question 6 — Explain out loud (6 marks)

In 3–5 sentences each, explain:

(a) Why a bi-elliptic transfer can beat a Hohmann transfer, and roughly at what radius ratio. (3)

(b) The physical mechanism of a gravity assist: why the spacecraft's speed changes in the Sun frame but its vv_\infty magnitude is unchanged in the planet frame. (3)


Answer keyMark scheme & solutions

Question 1 (12)

(a) ddt(r×r˙)=r˙×r˙+r×r¨=0+r×(μr3r)=0\dfrac{d}{dt}(\mathbf r\times\dot{\mathbf r}) = \dot{\mathbf r}\times\dot{\mathbf r} + \mathbf r\times\ddot{\mathbf r} = 0 + \mathbf r\times(-\tfrac{\mu}{r^3}\mathbf r) = 0. (1) for product rule, (1) first term zero, (1) second term zero because rr¨\mathbf r\parallel\ddot{\mathbf r}. Hence h\mathbf h constant → motion planar.

(b) Take r¨×h=μr3r×(r×r˙)\ddot{\mathbf r}\times\mathbf h = -\tfrac{\mu}{r^3}\mathbf r\times(\mathbf r\times\dot{\mathbf r}). Using BAC–CAB: r×(r×r˙)=r(rr˙)r˙(r2)\mathbf r\times(\mathbf r\times\dot{\mathbf r}) = \mathbf r(\mathbf r\cdot\dot{\mathbf r}) - \dot{\mathbf r}(r^2), and rr˙=rr˙\mathbf r\cdot\dot{\mathbf r}=r\dot r. (2) So RHS =μ(r˙r2rr˙r)=μddt ⁣(rr)= -\mu\left(\dfrac{\dot r}{r^2}\mathbf r - \dfrac{\dot{\mathbf r}}{r}\right) = \mu\dfrac{d}{dt}\!\left(\dfrac{\mathbf r}{r}\right). (2) LHS =ddt(r˙×h)=\dfrac{d}{dt}(\dot{\mathbf r}\times\mathbf h) since h\mathbf h const. Integrate: r˙×h=μ(rr+e)\dot{\mathbf r}\times\mathbf h = \mu\left(\dfrac{\mathbf r}{r}+\mathbf e\right), where e\mathbf e is the constant eccentricity vector. (1) Dot with r\mathbf r: r(r˙×h)=(r×r˙)h=h2\mathbf r\cdot(\dot{\mathbf r}\times\mathbf h)=(\mathbf r\times\dot{\mathbf r})\cdot\mathbf h = h^2. RHS =μ(r+re)=μ(r+recosθ)=\mu(r + \mathbf r\cdot\mathbf e)=\mu(r + re\cos\theta). Solve: (1) r=h2/μ1+ecosθ.r=\frac{h^2/\mu}{1+e\cos\theta}. e\mathbf e points toward perigee; its magnitude is the eccentricity. (1)

(c) e=0e=0 circle; 0<e<10<e<1 ellipse; e=1e=1 parabola; e>1e>1 hyperbola. (2)

Question 2 (10)

(a) At any point ε=12v2μr\varepsilon=\tfrac12 v^2-\tfrac{\mu}{r} (energy conservation). (1) Setting ε=μ2a\varepsilon=-\tfrac{\mu}{2a} (1): 12v2μr=μ2a\tfrac12 v^2-\tfrac{\mu}{r}=-\tfrac{\mu}{2a}. (1) Solve for v2v^2: v2=μ ⁣(2r1a)v^2=\mu\!\left(\tfrac{2}{r}-\tfrac{1}{a}\right). (1)

(b) Areal velocity dAdt=h2\dfrac{dA}{dt}=\dfrac{h}{2} (Kepler II, from h\mathbf h const). (1) Over one period area swept = ellipse area: πab=h2T\pi a b = \dfrac{h}{2}T. (1) So T=2πabhT=\dfrac{2\pi ab}{h}. With b=a1e2b=a\sqrt{1-e^2} and h2=μp=μa(1e2)h^2=\mu p=\mu a(1-e^2)h=μa(1e2)h=\sqrt{\mu a(1-e^2)}. (1) T=2πaa1e2μa(1e2)=2πa3/2μ.(2)T=\frac{2\pi a\cdot a\sqrt{1-e^2}}{\sqrt{\mu a(1-e^2)}}=\frac{2\pi a^{3/2}}{\sqrt\mu}. \textbf{(2)} Square: T2=4π2μa3T^2=\dfrac{4\pi^2}{\mu}a^3. (1)

Question 3 (12)

v1=μ/r1=3.986×105/6778=7.6686 km/sv_1=\sqrt{\mu/r_1}=\sqrt{3.986\times10^5/6778}=7.6686\ \text{km/s}. (1) v2=μ/r2=3.986×105/42164=3.0747 km/sv_2=\sqrt{\mu/r_2}=\sqrt{3.986\times10^5/42164}=3.0747\ \text{km/s}. (1)

(b) at=12(r1+r2)=12(48942)=24471 kma_t=\tfrac12(r_1+r_2)=\tfrac12(48942)=24471\ \text{km}. (1) Perigee speed vp=μ(2/r11/at)=3.986×105(2/67781/24471)v_p=\sqrt{\mu(2/r_1-1/a_t)}=\sqrt{3.986\times10^5(2/6778-1/24471)} =3.986×105(2.9508×1044.0865×105)=3.986×1052.5421×104=10.064 km/s=\sqrt{3.986\times10^5(2.9508\times10^{-4}-4.0865\times10^{-5})}=\sqrt{3.986\times10^5\cdot2.5421\times10^{-4}}=10.064\ \text{km/s}. (1.5) Apogee speed va=μ(2/r21/at)=3.986×105(4.7434×1054.0865×105)=3.986×1056.569×106=1.6183 km/sv_a=\sqrt{\mu(2/r_2-1/a_t)}=\sqrt{3.986\times10^5(4.7434\times10^{-5}-4.0865\times10^{-5})}=\sqrt{3.986\times10^5\cdot6.569\times10^{-6}}=1.6183\ \text{km/s}. (1.5)

(c) Δv1=vpv1=10.0647.6686=2.395 km/s\Delta v_1=v_p-v_1=10.064-7.6686=2.395\ \text{km/s}. (1.5) Δv2=v2va=3.07471.6183=1.456 km/s\Delta v_2=v_2-v_a=3.0747-1.6183=1.456\ \text{km/s}. (1.5) Total Δv=3.852 km/s\Delta v=3.852\ \text{km/s}. (1)

(d) t=πat3/μ=π244713/3.986×105=π3.674×107=π606119040 s5.29 ht=\pi\sqrt{a_t^3/\mu}=\pi\sqrt{24471^3/3.986\times10^5}=\pi\sqrt{3.674\times10^7}=\pi\cdot6061\approx19040\ \text{s}\approx5.29\ \text{h}. (2)

Question 4 (10)

(a) Mean anomaly M=n(ttp)M=n(t-t_p) increases uniformly, n=2π/Tn=2\pi/T. (1) Define eccentric anomaly EE via the auxiliary circle of radius aa. Using Kepler II, area swept from perigee is proportional to time: M2π=areaπa2\dfrac{M}{2\pi}=\dfrac{\text{area}}{\pi a^2}. (1) The area (sector on ellipse) computed from the circle geometry = 12a2(EesinE)\tfrac12 a^2(E-e\sin E). (1) Equating: M=EesinEM=E-e\sin E. (1)

(b) Pseudocode: (4, 1 each for guess, formula, loop, stop)

def kepler(M, e, tol=1e-10):
    E = M if e < 0.8 else 3.14159   # initial guess
    while True:
        dE = (E - e*sin(E) - M) / (1 - e*cos(E))
        E -= dE
        if abs(dE) < tol:
            return E

Iteration: En+1=EnEnesinEnM1ecosEnE_{n+1}=E_n-\dfrac{E_n-e\sin E_n-M}{1-e\cos E_n}.

(c) E0=0.8E_0=0.8. f=0.80.3sin0.80.8=0.3sin0.8=0.3(0.71736)=0.21521f=0.8-0.3\sin0.8-0.8=-0.3\sin0.8=-0.3(0.71736)=-0.21521. f=10.3cos0.8=10.3(0.69671)=0.79099f'=1-0.3\cos0.8=1-0.3(0.69671)=0.79099. (1) E1=0.8(0.21521/0.79099)=0.8+0.27208=1.0721 radE_1=0.8-(-0.21521/0.79099)=0.8+0.27208=1.0721\ \text{rad}. (1)

Question 5 (10)

(a) Formula stated. (1) The oblate bulge adds mass near the equator; the resulting non-central torque-free but non-Keplerian potential makes the orbital plane precess about Earth's spin axis, so Ω\Omega drifts. (2) At i=90°i=90° (polar), cosi=0\cos i=0 ⇒ symmetry makes net nodal torque zero, Ω˙=0\dot\Omega=0. (1)

(b) SSO requires Ω˙=+1.991×107 rad/s\dot\Omega = +1.991\times10^{-7}\ \text{rad/s} (matching Earth's mean motion about Sun, 2π/year2\pi/\text{year}), so the orbital plane keeps a fixed angle to the Sun. (2) Since Ω˙>0\dot\Omega>0 requires cosi<0\cos i<0, the orbit must be retrograde, i>90°i>90° (typically ~98°). (1)

(c) n=μ/a3=3.986×105/70783=3.986×105/3.546×1011=1.0603×103 rad/sn=\sqrt{\mu/a^3}=\sqrt{3.986\times10^5/7078^3}=\sqrt{3.986\times10^5/3.546\times10^{11}}=1.0603\times10^{-3}\ \text{rad/s}. (1) cosi=Ω˙32J2n(R/a)2\cos i = \dfrac{\dot\Omega}{-\tfrac32 J_2 n (R_\oplus/a)^2} (with e=0e=0). (R/a)2=(6378/7078)2=0.8119(R_\oplus/a)^2=(6378/7078)^2=0.8119. Denominator =32(1.0826×103)(1.0603×103)(0.8119)=1.3980×106=-\tfrac32(1.0826\times10^{-3})(1.0603\times10^{-3})(0.8119)=-1.3980\times10^{-6}. (1) cosi=1.991×107/(1.3980×106)=0.14242\cos i=1.991\times10^{-7}/(-1.3980\times10^{-6})=-0.14242i=arccos(0.14242)=98.19°i=\arccos(-0.14242)=98.19°. (1)

Question 6 (6)

(a) Bi-elliptic uses an intermediate apogee rbr2r_b\gg r_2; the plane-change/velocity-change burns happen where orbital speed is small, cutting the required Δv\Delta v despite three burns. It beats Hohmann when the ratio r2/r111.94r_2/r_1\gtrsim 11.94 (and always for >15.58>15.58), at the cost of much longer transfer time. (3)

(b) In the planet's frame the encounter is a hyperbola; energy is conserved, so incoming and outgoing vv_\infty have equal magnitude — only the direction rotates. In the Sun's frame you add the planet's heliocentric velocity to v\mathbf v_\infty; because the direction changed, the vector sum has a different magnitude, so the spacecraft gains (or loses) heliocentric speed while the planet loses (gains) a negligible amount. (3)

[
  {"claim":"Hohmann v1 circular LEO", "code":"mu=3.986e5; r1=6778.0; v1=sqrt(mu/r1); result = abs(float(v1)-7.6686)<0.01"},
  {"claim":"Hohmann total delta-v ~3.852 km/s", "code":"mu=3.986e5; r1=6778.0; r2=42164.0; at=(r1+r2)/2; v1=sqrt(mu/r1); v2=sqrt(mu/r2); vp=sqrt(mu*(2/r1-1/at)); va=sqrt(mu*(2/r2-1/at)); dv=(vp-v1)+(v2-va); result = abs(float(dv)-3.852)<0.01"},
  {"claim":"Transfer time ~19040 s", "code":"import sympy as sp; mu=3.986e5; at=24471.0; t=sp.pi*sqrt(at**3/mu); result = abs(float(t)-19040)<50"},
  {"claim":"Kepler one Newton step E1~1.0721", "code":"M=0.8; e=0.3; E0=0.8; f=E0-e*sin(E0)-M; fp=1-e*cos(E0); E1=E0-f/fp; result = abs(float(E1)-1.0721)<0.001"},
  {"claim":"SSO inclination ~98.19 deg", "code":"mu=3.986e5; a=7078.0; Re=6378.0; J2=1.0826e-3; Om=1.991e-7; n=sqrt(mu/a**3); c=Om/(-1.5*J2*n*(Re/a)**2); i