133 lines
4.8 KiB
Markdown
133 lines
4.8 KiB
Markdown
---
|
|
title: Orbit manoeuvres
|
|
tags: maneuvers, impulse, finite-burn, hohmann, delta-v
|
|
summary: How a change in orbit is made, what it costs in propellant, and why the cheapest transfer between two circular orbits is the one everybody uses.
|
|
---
|
|
|
|
# Orbit manoeuvres
|
|
|
|
A manoeuvre changes an orbit. The satellite is at one point on a closed curve and
|
|
must end up at another point on a different closed curve, and the change has to
|
|
be paid for in propellant. This is the subject: how large the change is, how
|
|
much it costs, and how the cost changes when the orbit is not circular.
|
|
|
|
## Definition
|
|
|
|
A **delta-v** is the magnitude of a change in velocity. A manoeuvre is planned
|
|
as a sequence of delta-v, and the propellant cost follows from their sum by the
|
|
rocket equation.
|
|
|
|
An **impulse** applies the whole change in zero time. It is the right model for
|
|
a burn that is short compared with the orbital period, which for a low Earth
|
|
orbit means anything under a minute.
|
|
|
|
A **finite burn** applies a thrust over a duration long enough that the
|
|
satellite moves while thrusting. The distinction is not academic: a low Earth
|
|
orbital transfer assisted by electric propulsion can last for hours and its
|
|
gravity loss is a substantial fraction of the ideal cost.
|
|
|
|
## Mathematics
|
|
|
|
### The rocket equation
|
|
|
|
The velocity a vehicle can produce depends on the exhaust velocity of its
|
|
propellant and on the mass ratio,
|
|
|
|
```text
|
|
delta_v = v_e * ln (m_0 / m_f)
|
|
```
|
|
|
|
with `v_e` the exhaust velocity. Since a logarithmic function grows slowly,
|
|
doubling the propellant gives far less than double the velocity. This is why
|
|
the delta-v of a mission is the primary constraint on its design, and why
|
|
staging exists.
|
|
|
|
### The Hohmann transfer
|
|
|
|
Between two coplanar circular orbits the cheapest two-impulse transfer is the
|
|
Hohmann transfer. The first burn raises or lowers the apoapsis to match the
|
|
target, the satellite coasts through half a revolution, and the second burn
|
|
circularises.
|
|
|
|
With `r_1`, `r_2` the two radii and `a_t = (r_1 + r_2) / 2` the transfer
|
|
semi-major axis, the two burns follow from the vis-viva equation,
|
|
|
|
```text
|
|
delta_v_1 = sqrt( mu (2/r_1 - 1/a_t) ) - sqrt( mu / r_1 )
|
|
delta_v_2 = sqrt( mu / r_2 ) - sqrt( mu (2/r_2 - 1/a_t) )
|
|
```
|
|
|
|
and the coast takes half the period of the transfer ellipse,
|
|
|
|
```text
|
|
t_coast = pi sqrt( a_t^3 / mu )
|
|
```
|
|
|
|
A raise from 300 km to geostationary costs about 4.1 km/s, which is a
|
|
large fraction of what a launch vehicle can provide, and most of it is the
|
|
first burn.
|
|
|
|
### Plane changes
|
|
|
|
A transfer that also changes inclination is more expensive than a coplanar
|
|
one. The combined manoeuvre, raising the apoapsis and rotating the plane at
|
|
the same time, costs more than doing either alone. The optimum for a given
|
|
radius change and a given inclination change is a two-burn sequence in which
|
|
the plane change is made at the smaller radius, where the velocity is lowest
|
|
and therefore the same angular rotation is cheapest in delta-v.
|
|
|
|
```text
|
|
delta_v_plane ~ V * 2 sin( inclination change / 2 )
|
|
```
|
|
|
|
with `V` the circular speed at the radius where the rotation is performed.
|
|
|
|
### Finite burns
|
|
|
|
Over a finite burn the satellite is still moving against gravity and the
|
|
thrust does not act along the instantaneous velocity. The achieved velocity
|
|
change is less than the ideal `a * t`, and the shortfall is the gravity loss,
|
|
|
|
```text
|
|
gravity loss = | delta_v_ideal | - | delta_v_achieved |
|
|
```
|
|
|
|
which grows with burn duration. A chemical burn is short enough to ignore it.
|
|
An electric thruster is not.
|
|
|
|
## Unit conventions
|
|
|
|
- Delta-v and velocity in kilometres per second.
|
|
- Radii in kilometres.
|
|
- The gravitational parameter `mu` in cubic kilometres per second squared.
|
|
- Inclinations in radians; degrees hide a factor of about 57 when used in a
|
|
sine.
|
|
|
|
## Applicability
|
|
|
|
The Hohmann transfer is optimal only for coplanar circular orbits, and only
|
|
among two-impulse transfers. Real missions are rarely that: the orbits are
|
|
elliptical, inclined, or both, and a bi-elliptic transfer can beat Hohmann
|
|
when the ratio of the two radii exceeds about 11.94, at the cost of a much
|
|
longer flight.
|
|
|
|
The impulsive model is adequate when the burn is a small fraction of the
|
|
orbital period. It is not adequate for a spiral orbit raise by electric
|
|
propulsion, where the "burn" is the whole mission.
|
|
|
|
A plane change is rarely worth it. Raising the inclination to reach a
|
|
different ground track costs delta-v proportional to the inclination and
|
|
often exceeds the entire launch budget, which is why planes are chosen at
|
|
launch and then maintained rather than changed.
|
|
|
|
## Standards
|
|
|
|
- Hohmann transfer: derived from the vis-viva equation; see Vallado,
|
|
*Fundamentals of Astrodynamics and Applications*
|
|
- Rocket equation: Tsiolkovsky, 1903
|
|
|
|
## See also
|
|
|
|
- [[keplerian-elements]] for the orbit being changed
|
|
- [[sgp4]] for propagating the result of a manoeuvre
|
|
- [[perturbations]] for what the propellant estimate omits |