3.2.23Orbital Mechanics & Astrodynamics

Combined maneuvers — optimal split between plane change and velocity change

1,811 words8 min readdifficulty · medium

WHAT are we combining?

Two velocity vectors that differ in both magnitude and direction:

  • v1v_1 = speed before the burn
  • v2v_2 = speed after the burn
  • θ\theta = angle between the two velocity vectors (the plane-change angle)

The maneuver we must supply is the vector Δv=v2v1\vec{\Delta v} = \vec v_2 - \vec v_1.


HOW to derive the combined Δv\Delta v (law of cosines)

Place v1\vec v_1 and v2\vec v_2 tail-to-tail with angle θ\theta between them.

Δv=v2v1\vec{\Delta v} = \vec v_2 - \vec v_1 Δv2=ΔvΔv=v222v1 ⁣ ⁣v2+v12|\Delta v|^2 = \vec{\Delta v}\cdot\vec{\Delta v} = v_2^2 - 2\,\vec v_1\!\cdot\!\vec v_2 + v_1^2

Since v1v2=v1v2cosθ\vec v_1\cdot\vec v_2 = v_1 v_2\cos\theta:

Sanity checks (Forecast-then-Verify):

  • θ=0\theta = 0 (no plane change): Δv=(v1v2)2=v2v1\Delta v = \sqrt{(v_1-v_2)^2}=|v_2-v_1|. ✅ pure speed change.
  • v1=v2=vv_1=v_2=v (pure plane change): Δv=v22cosθ=2vsin(θ/2)\Delta v = v\sqrt{2-2\cos\theta}=2v\sin(\theta/2). ✅ the classic plane-change formula.

HOW to split the plane change optimally

Consider the standard problem: a Hohmann transfer between two circular orbits where the final orbit is also inclined by total angle ii relative to the initial. You have two burns:

  • burn 1 (departure, at speed vpv_p — high, near the smaller orbit),
  • burn 2 (arrival, at speed vav_a — low, near the larger orbit).

Let the plane change ss be done at burn 1 and (is)(i-s) at burn 2. Each burn is a combined maneuver:

Δv1(s)=v1,i2+v1,f22v1,iv1,fcoss\Delta v_1(s) = \sqrt{v_{1,i}^2 + v_{1,f}^2 - 2 v_{1,i}v_{1,f}\cos s} Δv2(s)=v2,i2+v2,f22v2,iv2,fcos(is)\Delta v_2(s) = \sqrt{v_{2,i}^2 + v_{2,f}^2 - 2 v_{2,i}v_{2,f}\cos(i-s)}

Total Δvtot(s)=Δv1(s)+Δv2(s)\Delta v_{\text{tot}}(s) = \Delta v_1(s) + \Delta v_2(s). Minimize:

dΔvtotds=0\frac{d\,\Delta v_{\text{tot}}}{ds} = 0

Figure — Combined maneuvers — optimal split between plane change and velocity change

Worked Example 1 — LEO→GEO with inclination change

Given: transfer between r1r_1 (LEO, fast) and r2r_2 (GEO, slow), total inclination change i=28.5i=28.5^\circ (Cape Canaveral to equatorial GEO).

Approximate speeds (typical): perigee burn between v1,i=7.73v_{1,i}=7.73 and v1,f=10.15v_{1,f}=10.15 km/s; apogee burn between v2,i=1.61v_{2,i}=1.61 and v2,f=3.07v_{2,f}=3.07 km/s.

Step — do all plane change at apogee (s=0s=0): Δv2=1.612+3.0722(1.61)(3.07)cos28.5=1.83 km/s\Delta v_2 = \sqrt{1.61^2+3.07^2-2(1.61)(3.07)\cos 28.5^\circ}=1.83\text{ km/s} Why this step? Apogee has the smallest velocities, so 2vsin2v\sin costs least here.

Step — split optimally: numerically minimizing Δvtot(s)\Delta v_{\text{tot}}(s) gives s2.2s\approx 2.2^\circ at perigee, 26.326.3^\circ at apogee. Why this step? The equal-marginal-cost condition; a tiny share at perigee shaves a little more.

Step — compare with the naive "all at perigee": Δv1(all 28.5)=7.732+10.1522(7.73)(10.15)cos28.5=5.0 km/s (!)\Delta v_1(\text{all }28.5^\circ)=\sqrt{7.73^2+10.15^2-2(7.73)(10.15)\cos28.5^\circ}=5.0\text{ km/s (!)} Why this step? Shows the penalty: doing plane change at the fast burn nearly doubles the whole mission cost.

Takeaway: save the plane change for apoapsis, keep only a whisker at perigee.


Worked Example 2 — Pure equal-speed plane change vs split

Suppose v1=v2=vv_1=v_2=v (same circular orbit, only tilt it by θ\theta).

Single burn: Δv=2vsin(θ/2)\Delta v = 2v\sin(\theta/2).

Bi-elliptic trick (conceptual): raise apoapsis far out (cheap prograde burn), do the plane change there where vv is tiny (so 2vsin(θ/2)2v\sin(\theta/2) is tiny), then drop back. Why this step? Splitting velocity change (raise/lower) from where the plane change happens exploits low speed — for large θ\theta this beats the single burn. Same principle as combined-maneuver splitting: put the turn where you're slow.



Recall Feynman: explain to a 12-year-old

Imagine running fast and wanting to turn to face a new direction. If you turn while sprinting, you skid hard — it takes tons of effort. If you slow down first, then turn, then speed up, turning is easy. In space, "turning" (changing your orbit's tilt) costs the most when you're moving fast. So spacecraft cleverly do their turning at the far, slow part of the orbit, and let a single angled push do the speeding-up and the turning together — like one smooth swerve instead of two jerks.


Flashcards

Combined-maneuver Δv\Delta v formula
Δv=v12+v222v1v2cosθ\Delta v=\sqrt{v_1^2+v_2^2-2v_1v_2\cos\theta}
Why is it the law of cosines?
Δv\Delta v is the third side of the velocity triangle with included angle θ\theta between v1v_1 and v2v_2.
Pure plane-change cost (equal speeds)
Δv=2vsin(θ/2)\Delta v = 2v\sin(\theta/2).
Cost of a 6060^\circ plane change at speed vv
Δv=v\Delta v = v (the whole orbital speed again).
Where should most of the plane change be done, and why?
At apoapsis / the slower burn, because cost scales with the velocity there.
Optimal-split condition (words)
The marginal Δv\Delta v cost of one more degree of plane change is equal at both burns.
Optimal-split condition (formula)
v1,iv1,fsinsΔv1=v2,iv2,fsin(is)Δv2\dfrac{v_{1,i}v_{1,f}\sin s}{\Delta v_1}=\dfrac{v_{2,i}v_{2,f}\sin(i-s)}{\Delta v_2}.
Why combine instead of two separate burns?
Triangle inequality: v12+v222v1v2cosθv2v1+2v2sin(θ/2)\sqrt{v_1^2+v_2^2-2v_1v_2\cos\theta}\le |v_2-v_1|+2v_2\sin(\theta/2).
Sanity check θ=0\theta=0
Δv=v2v1\Delta v=|v_2-v_1| (pure speed change).
For LEO→GEO from 28.5°, roughly how is the split?
Nearly all plane change at apogee (~26°), only ~2° at perigee.

Connections

Concept Map

is expensive when

combine into

cheaper than

dv is third side of

solved by

gives

theta=0 limit

v1=v2 limit

shows why avoid

applied per burn

minimize total dv

put most where

Plane change fights orbital speed

Speed v is large

Need speed + direction change

Single angled dog-leg burn

Two separate burns

Velocity triangle v1 v2 theta

Law of cosines

dv = sqrt v1^2+v2^2-2v1v2 cos theta

Pure speed change

2v sin theta/2

Hohmann transfer with inclination i

Optimal split of plane change s

Speed is slowest at apoapsis

Hinglish (regional understanding)

Intuition Hinglish mein samjho

Dekho, orbital mechanics mein sabse mehenga kaam hota hai plane change — yaani orbit ka tilt (inclination) badalna. Kyunki jab tum tez chal rahe ho, us velocity ko sideways modna padta hai, aur uska cost 2vsin(θ/2)2v\sin(\theta/2) hota hai. Matlab agar θ=60\theta=60^\circ hai to poora orbital speed dubara kharch ho jaata hai! Isliye rule simple hai: "jahan slow ho, wahan turn karo." Apoapsis pe speed sabse kam hoti hai, to plane change wahan sasta padta hai.

Ab agar tumhein speed bhi change karni hai aur plane bhi (jaise Hohmann transfer mein), to do alag burns karne ke bajaye ek hi tirchha (angled) burn karo — dono kaam ek saath. Iska cost law of cosines se aata hai: Δv=v12+v222v1v2cosθ\Delta v=\sqrt{v_1^2+v_2^2-2v_1v_2\cos\theta}. Ye triangle inequality ki wajah se hamesha do alag burns se kam ya barabar hota hai. Yahi "combined maneuver" ka jaadu hai.

Optimal split ka funda: total plane change ko dono burns ke beech itne ratio mein baanto ki har burn ka marginal cost (ek extra degree ka kharcha) equal ho jaaye. Jo burn slow hai (apogee), wahan turning sasti hai, isliye zyaadatar plane change wahan chala jaata hai — LEO se GEO waale case mein perigee pe sirf ~2° bachta hai, baaki ~26° apogee pe. Exam aur real mission dono mein: turn where slow, burn in one throw!

Go deeper — visual, from zero

Test yourself — Orbital Mechanics & Astrodynamics

Connections