Wakker (2015): Fundamentals of Astrodynamics

Source

Wakker, Karel F., “Fundamentals of Astrodynamics,” Faculty of Aerospace Engineering, Delft University of Technology, January 2015. ISBN 978-94-6186-419-2. 690 pages. Open Access via the TU Delft Institutional Repository.

Download PDF · TU Delft Repository

Supplementary Reference

This is NOT a direct SGP4 lineage document. Wakker does not describe or implement SGP4, SDP4, or any of the Spacetrack models. This page documents the textbook as a pedagogical companion — it provides the mathematical foundations that SGP4 builds upon, derived from first principles with every intermediate step shown.

Why This Matters for SGP4

The primary SGP4 papers — Brouwer (1959), Hoots (1981), STR#3 — are terse. They assume the reader already knows Lagrange’s planetary equations, the method of averaging, Delaunay canonical variables, and the structure of the geopotential. Wakker fills that gap.

Across 184 pages of perturbation theory (Chapters 20-23), Wakker derives from first principles every piece of mathematical machinery that SGP4 implements:

• Lagrange’s planetary equations — the six differential equations that govern how orbital elements change under any perturbation. Brouwer’s entire paper is a specific application of these equations to the $J_2$-$J_5$ geopotential.
• J2 secular rates — the constant-rate changes in $\Omega$, $\omega$, and $M$ that appear directly in sgp4.f lines 29-52. Wakker shows how they emerge from averaging the Lagrange equations over an orbit.
• Mean element theory — the separation of perturbations into secular, long-period, and short-period components, and the distinction between osculating and mean elements that defines what a TLE actually encodes.
• Kaula’s spectral decomposition — the frequency analysis that explains orbital resonance, which is the physical basis for SDP4’s DEEP subroutine.
• The critical inclination singularity — why Brouwer’s long-period terms blow up near $i = 63.4\degree$ and what the singularity means physically.

For anyone trying to understand why SGP4 works the way it does rather than just how to run it, Wakker is the missing textbook.

Lagrange’s Planetary Equations

Chapter 22 provides the most thorough single-textbook derivation of Lagrange’s planetary equations available — 28 pages from Newton’s laws to the complete set of six differential equations. The derivation builds through Lagrange brackets, canonical (Delaunay) variables, and the Gauss form, showing every intermediate step.

The core result (eq. 22.35):

$\frac{da}{dt} = \frac{2}{na}\frac{\partial \mathcal{R}}{\partial M}$

$\frac{de}{dt} = \frac{1-e^2}{na^2 e}\frac{\partial \mathcal{R}}{\partial M} - \frac{\sqrt{1-e^2}}{na^2 e}\frac{\partial \mathcal{R}}{\partial \omega}$

$\frac{di}{dt} = \frac{1}{na^2\sqrt{1-e^2}\sin i}\left[\cos i \frac{\partial \mathcal{R}}{\partial \omega} - \frac{\partial \mathcal{R}}{\partial \Omega}\right]$

where $\mathcal{R}$ is the disturbing function (the non-Keplerian part of the potential) and $n$, $a$, $e$, $i$, $M$, $\omega$, $\Omega$ are the standard Keplerian elements.

Three additional equations for $d\Omega/dt$, $d\omega/dt$, and $dM/dt$ complete the set. Note the singularities at $e = 0$ (division by $e$ in the eccentricity equation) and $i = 0$ (division by $\sin i$ in the inclination equation). These coordinate singularities are the reason Lyddane (1963) had to reformulate Brouwer’s theory using Poincare variables before SGP4 could handle near-circular and near-equatorial orbits.

The Delaunay canonical form (Section 22.2), using variables $L$, $G$, $H$ and their conjugate angles $l$, $g$, $h$, is the formalism that Brouwer actually used. Wakker shows that the Delaunay variables are just a convenient repackaging of the Keplerian elements:

$L = \sqrt{\mu a}, \quad G = \sqrt{\mu a(1-e^2)}, \quad H = G\cos i$

This form makes Hamilton’s equations directly applicable and is the starting point for both Brouwer’s von Zeipel method and Kozai’s method of averaging.

J2 Secular Rates

Section 23.3 derives the first-order secular rates of change due to the $J_2$ harmonic of the geopotential. These are the rates that appear directly in the SGP4 FORTRAN code (sgp4.f lines 29-52):

$\dot{\Omega}_s = -\frac{3}{2}\frac{\bar{n} J_2 R^2}{p^2}\cos i$

$\dot{\omega}_s = \frac{3}{2}\frac{\bar{n} J_2 R^2}{p^2}\left(\frac{5}{2}\cos^2 i - 1\right)$

$\dot{M}_s = \bar{n}\left[1 + \frac{3}{2}\frac{J_2 R^2}{p^2}\sqrt{1-e^2}\left(1 - \frac{3}{2}\sin^2 i\right)\right]$

where $p = a(1-e^2)$ is the semi-latus rectum and $\bar{n}$ is the perturbed mean motion.

The method is straightforward: substitute the $J_2$ disturbing function into Lagrange’s equations, average over the mean anomaly $M$ (eliminating short-period terms), then average over the argument of perigee $\omega$ (eliminating long-period terms). What remains varies linearly with time — the secular rates.

Symbol Mapping: Wakker to SGP4

Wakker Symbol	SGP4/STR#3 Variable	Meaning
$a$	A0 (or AODP)	Mean semi-major axis
$e$	E0 (or ECCO)	Mean eccentricity
$i$	I0 (or INCLO)	Mean inclination
$n$	XN0 (or XNO)	Mean motion (Keplerian)
$\bar{n}$	XNODP	Perturbed mean motion
$J_2$	CK2 (= $\frac{1}{2} J_2 R^2$)	J2 coefficient (absorbed into CK2)
$p$	A0DP * (1-E0^2)	Semi-latus rectum
$\dot{\Omega}$	XNODOT	Secular rate of RAAN
$\dot{\omega}$	OMGDOT	Secular rate of argument of perigee
$\dot{M}$	XMDOT	Secular rate of mean anomaly

The Critical Inclination

The secular rate of the argument of perigee vanishes when:

$\frac{5}{2}\cos^2 i - 1 = 0 \quad \Rightarrow \quad \cos^2 i = \frac{2}{5} \quad \Rightarrow \quad i = 63.435\degree \text{ or } 116.565\degree$

At this inclination, perigee freezes — it does not precess around the orbit. This is the physical basis of Molniya orbits (which keep apogee over high northern latitudes) and the mathematical source of the $(1 - 5\cos^2 i)^{-1}$ singularity in Brouwer’s long-period terms.

Wakker notes (p. 626) that Deprit showed in 1983 that the critical inclination is not a physical singularity but an unstable equilibrium in phase space — the “troubles” are artifacts of the perturbation theory’s coordinate choice. SGP4 handles this by switching to the Lyddane variable set, which avoids division by the problematic factor.

Kozai Mean Elements

Section 23.3 (eq. 23.30-23.31) presents Kozai’s modified mean semi-major axis:

$\tilde{a} = a\left[1 + \frac{J_2 R^2}{2a^2(1-e^2)^{3/2}}\left(1 - \frac{3}{2}\sin^2 i\right)\right]$

where $\tilde{a}$ closely approximates the time-averaged geocentric distance.

This definition differs from Brouwer’s mean semi-major axis by 5-10 km for typical LEO orbits (equatorial vs. polar). The distinction matters operationally because TLE mean elements are defined relative to a specific mean-element theory:

• SGP4 uses Brouwer mean elements — both short-period and long-period terms have been removed
• SGP used Kozai mean elements — only short-period terms have been removed

Feeding Kozai-mean elements into a Brouwer-mean propagator (or vice versa) introduces systematic position error because the perturbation corrections no longer match what was removed during element fitting. The SGP4 initialization code handles the conversion between the two definitions internally, applying drag in the Kozai system before converting to Brouwer for gravitational perturbation calculations.

Kaula’s Spectral Theory

Section 23.5 presents Kaula’s spectral decomposition of the gravity field perturbations, expressing the disturbing function in terms of orbital elements:

$\mathcal{R} = \frac{\mu}{a}\sum_{n=2}^{\infty}\sum_{m=0}^{n}\sum_{p=0}^{n}\sum_{q=-\infty}^{\infty}\left(\frac{R}{a}\right)^n F_{n,m,p}(i)\, G_{n,p,q}(e)\, S_{n,m,p,q}(\omega, M, \Omega, \theta_G)$

where $F_{n,m,p}(i)$ and $G_{n,p,q}(e)$ are the inclination and eccentricity functions, $\theta_G$ is the Greenwich sidereal time, and the four indices $(n, m, p, q)$ identify each spectral component.

The frequency of each component is:

$\dot{\Psi}_{n,m,p,q} = (n-2p)\dot{\omega} + (n-2p+q)\dot{M} + m(\dot{\Omega} - \dot{\theta}_G)$

When $\dot{\Psi} \to 0$, orbital resonance occurs. The perturbation amplitude grows without bound in the linear theory, and numerical integration becomes necessary.

This is the theoretical basis for SDP4’s DEEP subroutine. For 12-hour resonant orbits (GPS-type, where $\dot{M} \approx 2\dot{\theta}_G$) and 24-hour resonant orbits (GEO, where $\dot{M} \approx \dot{\theta}_G$), the frequency condition is nearly satisfied, and STR#1’s numerical integrator handles the perturbations that the analytical theory cannot.

Kaula’s framework also provides the taxonomy of perturbation types:

Type	Indices	Period	Example
Secular	$m=0$, $n-2p=0$, $q=0$	Infinite (constant rate)	$J_2$ precession of $\Omega$
Long-period	$m=0$, $q=\pm1$	Months to years	$J_3$ pear-shape effect
Short-period	$m=0$, $q\ne0$	~1 orbital period	$J_2$ radial oscillation
m-daily	$m\ne0$	~1/$m$ day	Tesseral harmonics
Resonance	$\dot{\Psi}\to0$	Weeks to years	GEO longitude drift

TEME Not Covered

Wakker does NOT cover the TEME (True Equator, Mean Equinox) reference frame. His treatment of precession and nutation uses the full IAU 2000A model (1,365 nutation terms), not the 4-term subset that SGP4 employs. Only Vallado’s Rev-1 paper formally defines TEME and provides the transformation matrices for converting SGP4 output to standard frames like J2000 and ITRF.

Chapter-Level SGP4 Relevance

Of Wakker’s 690 pages, approximately 184 (27%) contain SGP4-relevant content:

Chapter	Topic	Pages	SGP4 Connection
11	Reference frames and coordinates	243-284	Equatorial/ecliptic frames, precession/nutation, ICRS/GCRF
20	Perturbing forces and methods	527-554	$J_2$ geopotential, atmospheric drag, Cowell/Encke/variation-of-elements
21	Elementary perturbation analysis	555-584	$J_2$ effects, drag decay, luni-solar perturbations
22	Lagrange’s planetary equations	585-612	Full derivation, Delaunay canonical form, Gauss form, singularities
23	Gravity field perturbations	613-668	Secular/periodic variations, Kozai, Kaula, resonance, frozen orbits

The remaining 500 pages cover Keplerian orbits, the restricted three-body problem, orbital maneuvers, and interplanetary trajectories — useful astrodynamics context but not directly relevant to the SGP4 propagation chain.

---

Brouwer (1959): Mean Element Theory Applied these planetary equations to artificial satellites

Kozai (1959): Method of Averaging The alternative mean element theory Wakker compares

STR#1: Deep-Space Extensions Deep-space resonance handled by DEEP subroutine

Vallado (2013): Astrodynamics Textbook The other comprehensive textbook, focused on SGP4 implementation