# Zhang, Xu, Zhou (2017): Theory and Design Methods of Special Space Orbits **Yasheng Zhang, Yanli Xu, Haijun Zhou.** *Theory and Design Methods of Special Space Orbits.* Springer, 2017. 248 pages. A graduate-level Chinese astrodynamics textbook (translated to English) covering hovering orbits, spiral cruising orbits, multi-target rendezvous, initiative approaching orbits, fast responsive orbits, and Earth pole-sitter orbits. Each chapter is structured around a "Dynamic Model" derivation. Despite the consistent use of the word "dynamic," the entire book operates in acceleration units, with mass cancelled before propagation and reintroduced only at the very end as a multiplicative scaling for fuel and force budgets. Chapter 3 contains a particularly clean self-incrimination: a worked example where the authors explicitly state that applying thrust is mathematically equivalent to "reducing a part of earth's mass," producing what they call an "artificial sub-keplerian orbit." This source provides a third independent confirmation (after [[Artemis_II_Trajectory]] and [[Textbook_Orbital_Dynamics]]) that engineering astrodynamics is a kinematic discipline wearing dynamical vocabulary. It also extends the argument into spacecraft control, where active thrust would naively seem to require true dynamics. Parent note: [[Cosmological_Dynamics_Null]] First principles: [[First_Principles_Dynamics]] Companion sources: [[Textbook_Orbital_Dynamics]], [[Artemis_II_Trajectory]], [[2013_Bouw_Geocentricity_Forces_and_First_Principles]] --- ## Chapter 2: The master equation of motion ### Section 2.3.3 (p.34): Rectangular Coordinate Component ![[Zhang2017_SpecialOrbits-01_Ch2_Eq2.5_master_eqn.png]] The book sets out its **master equation of motion** for any spacecraft under orbit control as Eq. 2.5: $\ddot{\vec{r}} = -\frac{\mu}{r^3}\vec{r} + \vec{a} + \vec{a}_T \tag{2.5}$ The authors define the variables explicitly: > "Here, $\vec{r}$, $\mu$, $\vec{a}$ and $\vec{a}_T$ refer to the geocentric distance vector, the Earth's gravitational constant ($\mu = 3.986005 \times 10^{14}$ m³/s²), the perturbation acceleration, and the controlling acceleration, respectively." (p.34) > [!critical] The book's master equation has no mass and no force > Eq. 2.5 is the equation that governs every trajectory in the entire 248 page book. The left side is acceleration. The right side is one kinematic ratio ($\mu = 4\pi^2 a^3/T^2$ from satellite tracking, see [[Artemis_II_Trajectory]] §2) divided by a geometric quantity, plus two acceleration vectors. There is no mass anywhere. There is no force anywhere. The "controlling acceleration" $\vec{a}_T$ is the **thrust as an acceleration**, not as a force. Force only appears later, when an arbitrary spacecraft mass is multiplied in to compute fuel budgets. This is the entire physics of orbit control as the textbook frames it. Mass never enters the propagation. The dynamical content of the book is the choice of $\mu$ and the prescribed acceleration profile. Both are kinematic. The book is, by its own equation, a kinematic textbook. --- ## Chapter 3: Hovering Orbit (the smoking gun) Chapter 3 designs an orbit where one spacecraft "hovers" above another by applying continuous thrust to maintain a constant relative position. This is the most physically interesting case for the dynamics-vs-kinematics question, because hovering requires active thrust, and naively that should require knowing the spacecraft's mass to compute the force. The authors derive the answer two independent ways. Both come out mass-free. ### Section 3.2.1 (p.41): Mechanical analysis derivation ![[Zhang2017_SpecialOrbits-02_Ch3_eqs3.7-3.10.png]] Decomposing the hovering spacecraft's velocity into radial and circumferential components (Eq. 3.7): $V_r = \sqrt{\frac{\mu}{a(1-e^2)}} e \sin f, \qquad V_\phi = \sqrt{\frac{\mu}{a(1-e^2)}}(1 + e\cos f)$ Substituting the matching condition (Eq. 3.8) and computing the gravitational acceleration as the sum of the radial acceleration and the centripetal acceleration of circumferential rotation: $A_{Sr} = \dot{V}_{Sr} = \mu\frac{a_S}{a_T r_T^2} e\cos f \tag{3.9}$ $A_{S\phi} = \frac{V_{S\phi}^2}{r_S} = \mu\frac{a_S}{a_T r_T^2}(1 + e\cos f) \tag{3.10}$ Note: every right-hand side is $\mu$ times a ratio of orbital semi-major axes and a trigonometric function of the true anomaly. **No mass.** No force. ### Section 3.2.1 (p.42): The thrust acceleration formula (Eq. 3.11) ![[Zhang2017_SpecialOrbits-03_Ch3_eqs3.11-3.15.png]] The required thrust acceleration to maintain the hover, in radial-positive convention, is the difference between the Earth's gravitational acceleration $A_E = \mu/r_S^2$ and the orbital accelerations: $A = A_E - A_{S\phi} + A_{Sr} = \frac{\mu}{r_S^2} - \frac{\mu a_S}{a_T r_T^2}(1 + e\cos f) + \frac{\mu a_S}{a_T r_T^2} e\cos f$ $\boxed{A = \frac{\mu}{r_S^2}\left(1 - \frac{a_S^3}{a_T^3}\right)} \tag{3.11}$ > [!critical] The thrust acceleration depends only on $\mu$ and a ratio of cubed semi-major axes > This is the design equation for a hovering orbit. To make a spacecraft hover above another spacecraft, you need to apply a thrust acceleration $A$ given by **one kinematic ratio** ($\mu$, which is $4\pi^2 a^3/T^2$ from GRACE tracking) **times another kinematic ratio** ($1 - a_S^3/a_T^3$, the cube ratio of two orbital semi-major axes). There is no mass on either side. There is no force on either side. The hovering condition is a purely kinematic specification. ### Section 3.2.1 (p.42): The mass cancels before the equation propagates (Eq. 3.12-3.15) The same page presents the "two-body problem" derivation. The basic equation of motion is Eq. 3.12: $\ddot{\vec{r}} + \frac{\mu}{r^3}\vec{r} = 0 \tag{3.12}$ Reduced to the radial component: $\ddot{r} + \frac{\mu}{r^2} = 0 \tag{3.14}$ When an external force $F$ is applied (Eq. 3.15): $\ddot{r} + \frac{\mu}{r^2} = +A, \qquad \text{where } A = F/m_S$ > [!critical] The mass is divided out before the equation is even propagated > Equation 3.15 is exactly the Bouw move from [[2013_Bouw_Geocentricity_Forces_and_First_Principles]]. You start with $F = ma$, immediately rewrite as $A = F/m$, and the equation that gets propagated forward in time is $\ddot{r} + \mu/r^2 = A$. The mass $m_S$ that appeared in the definition of $A$ is gone from the propagation equation. It will not reappear until the very end of the chapter, when an arbitrary 1000 kg is multiplied back in to convert the kinematic answer into a force in newtons. The acceleration is the engineering specification. The force is a derivative quantity that depends on which spacecraft you happen to be flying. ### Section 3.2.1 (p.43): The "modified gravitational parameter" hides the thrust as a mass change ![[Zhang2017_SpecialOrbits-04_Ch3_eqs3.16-3.21.png]] The authors introduce a "modified $\muquot;: $\mu' = \mu - A r^2$ so that the equation of motion becomes: $\ddot{r} + \frac{\mu'}{r^2} = 0 \tag{3.16}$ The spacecraft now obeys an effective two body problem with a smaller central body. Working through the algebra (Eqs. 3.17-3.19), they obtain: $\mu' = \mu \frac{a_S^3}{a_T^3}$ Substituting back into $\mu' = \mu - A r^2$ recovers the same answer as the mechanical method: $A = \frac{\mu}{r_S^2}\left(1 - \frac{a_S^3}{a_T^3}\right) \tag{3.20}$ The authors then state explicitly: > "Apparently, the result obtained by adopting approximately the two-body problem method is the same as the one obtained by using the mechanical analysis method in physics. Thus the two methods have mutually authenticated the correctness of their derivation processes." (p.43) And finally Eq. 3.21: $F = m_S A \tag{3.21}$ > [!critical] Force is added at the end as a unit conversion > Eq. 3.21 is the only place in the entire derivation where the spacecraft mass enters. It enters by being multiplied into the kinematic answer $A$ to produce a force $F$. **This is the dynamics = kinematics × ($m/m$) trick from Bouw, applied to the engineering of a real spacecraft control problem.** The kinematic equation $A = (\mu/r_S^2)(1 - a_S^3/a_T^3)$ is complete. The force equation $F = m_S A$ adds nothing physical, it just converts the answer into different units. A spacecraft ten times heavier needs ten times the force to hover at the same place, but the *acceleration* required is identical. Acceleration is the physics. Force is the bookkeeping. ### Section 3.2.1.3 (p.44): The smoking gun quote ![[Zhang2017_SpecialOrbits-05_Ch3_subkeplerian_quote.png]] This is the page that should be tattooed inside the cover of every astrodynamics textbook. The authors interpret the "modified $\muquot; formulation as follows: > "The difference from the Third Law of Kepler is that the direct proportional coefficient of the square of the spacecraft's operation period to the cubic of the semi-major axis of its elliptical orbit is not earth's gravity $\mu$, but $\mu' = \mu - Ar^2$. **It can be interpreted as: the thrust imposed counteracts a part of earth's gravity, or 'reduces a part of earth's mass'. This kind of orbit can be described as an artificial 'sub-keplerian orbit'.**" (p.44, emphasis added) > [!critical] You can "reduce earth's mass" with a thrust because mass is a kinematic label > This is the cleanest demonstration in the engineering literature that mass is a label, not a substance. A thrust applied to a spacecraft is mathematically equivalent to reducing the Earth's mass. Why? Because the equation of motion contains $\mu = GM_\oplus$, and the only thing that the rest of the trajectory depends on is the *value* of $\mu$, not the source of that value. If you change $\mu$ by adding a thrust acceleration $-Ar^2$, you get an "artificial sub-keplerian orbit" that obeys Kepler's laws with a smaller effective $\mu$. The orbit cannot tell whether the smaller $\mu$ is "really" from a smaller Earth or "really" from an applied thrust. **The Earth's mass is whatever the kinematics say it is.** This is the same conclusion Bouw reaches in [[2013_Bouw_Geocentricity_Forces_and_First_Principles]] from first principles, and it is the same conclusion the [[Cosmological_Dynamics_Null]] reaches from spreadsheet analysis of ten orbiting bodies. Here it appears in a 2017 Springer engineering textbook as a casual aside. The phrase "artificial sub-keplerian orbit" is an accidental admission that *all* Keplerian orbits are kinematic configurations characterized by a single ratio $\mu$, and that ratio is fungible with thrust. The astronomical $\mu$ is the value such that no thrust is needed. The "artificial" $\mu$ is the value such that some thrust is needed. There is no physical distinction between them. Both are numbers in an equation. ### Section 3.2.1.3 (p.45): The numerical example exposes the trick ![[Zhang2017_SpecialOrbits-06_Ch3_numerical_example.png]] The authors work through a numerical example for a hovering spacecraft above a sun-synchronous target satellite at orbital radius $6.9 \times 10^5$ m, with a hovering distance of $2.0 \times 10^4$ m: > "We assume that the mass of the hovering spacecraft is 1000 kg... the required continuous thrust acceleration is $A = 0.067921$ m/s², and the continuous thrust $F = 67.921$ N." (p.45) > [!critical] The mass of 1000 kg is an assumption, the acceleration is the answer > Read the order of operations carefully. They **assume** the spacecraft mass is 1000 kg. Then they **compute** the thrust acceleration $A = 0.067921$ m/s² from Eq. 3.11. Then they **multiply** to get $F = 67.921$ N. The 1000 kg is an arbitrary input parameter, not a physical constraint. If they had assumed 500 kg, $A$ would be unchanged and $F$ would be 33.9605 N. If they had assumed 2000 kg, $A$ would be unchanged and $F$ would be 135.842 N. The acceleration is the physics. The force is whatever you choose to call physics. > > Compare to [[Artemis_II_Trajectory]] §6: every Artemis II $\Delta V$ is a difference between two velocities, computed without reference to the spacecraft's mass. The actual mass of Orion (about 26,500 kg of crew module plus service module) is irrelevant to the trajectory shape. It only matters for fuel mass and burn duration, which are downstream of the kinematic trajectory. --- ## Chapter 4: Spiral Cruising Orbit (Hill equation, no mass) ### Section 4.3.1 (p.104): Hill equation cruising trajectory ![[Zhang2017_SpecialOrbits-07_Ch4_Hill_eqs4.34-4.39.png]] The Hill (also called Clohessy-Wiltshire) equations describe the relative motion of two spacecraft in nearby orbits. The textbook gives the cruising trajectory as Eq. 4.34: $\frac{(x - x_{c0})^2}{b^2} + \frac{(y - y_{c0} + 1.5 x_{c0} n t)^2}{(2b)^2} = 1 \tag{4.34}$ The drift distance per period (Eq. 4.35), cruising velocity (Eq. 4.36), cruising radius (Eq. 4.38), and traversal period (Eq. 4.39): $L = 3\pi |x_{c0}|, \qquad V = 1.5 x_{c0} n, \qquad T \approx 2\pi a_T / V$ > [!note] The Hill equation contains no mass anywhere > The cruising orbit design uses positions ($x_{c0}, y_{c0}$), the mean motion $n$ of the target orbit (which is $2\pi/T_{\text{target}}$, kinematic), and a geometric parameter $b$. There are no masses, no forces, no $\mu$, no $G$, no $M$. The Hill equation is the local linearization of the two body problem in a rotating frame, and the linearization eliminates $\mu$ along with the mass. What remains is pure relative kinematics. ### Section 4.3.2 (p.106): E/I vector method ![[Zhang2017_SpecialOrbits-08_Ch4_EI_vector_eqs4.43-4.44.png]] The authors give an alternative parameterization in terms of "configuration geometrical parameters" $(\Delta a, l, p, u, s, h)$: $\Delta a = a_2 - a_1, \quad p = a_1 |\Delta e|, \quad u = \arctan(\Delta e_y, \Delta e_x), \quad l = a_1(\Delta u + \Delta\Omega \cos i_1), \quad s = a_1 |\Delta i|, \quad h = \arctan(\Delta i_y, \Delta i_x)$ > [!note] Six parameters, all kinematic > The "configuration geometrical parameters" are: relative semi-major axis (length), in plane configuration size (length), in plane phase (angle), drift distance (length), out of plane size (length), out of plane phase (angle). Lengths and angles. No masses, no forces. The cruising orbit is fully specified by six geometric numbers. --- ## Chapter 8: Earth Pole-Sitter Orbit (the CR3BP "mass" parameter) ### Section 8.2.1 (p.226): The circular restricted three body problem ![[Zhang2017_SpecialOrbits-09_Ch8_CR3BP_eq8.1.png]] The chapter sets up the rotating frame equation of motion in the circular restricted three body problem (CR3BP) as Eq. 8.1: $\ddot{\vec{r}} + 2\vec{\omega}\times\dot{\vec{r}} + \vec{\omega}\times(\vec{\omega}\times\vec{r}) = -\nabla V + \vec{a} \tag{8.1}$ The left side is acceleration plus Coriolis plus centrifugal. The right side is the gradient of a potential plus a control acceleration. **Still no mass on either side.** The Coriolis and centrifugal terms are exactly the kinematic identities Bouw derives in Appendix E (see [[2013_Bouw_Geocentricity_Forces_and_First_Principles]]). ### Section 8.2.1 (p.227): The "mass parameter" is a kinematic ratio ![[Zhang2017_SpecialOrbits-10_Ch8_mass_parameter_eq8.2.png]] The gravitational potential is given as Eq. 8.2: $V(\vec{r}) = -\left(\frac{1-\mu}{r_1} + \frac{\mu}{r_2}\right), \qquad \mu = \frac{m_2}{m_1 + m_2} \tag{8.2}$ with $G m_1 = 1 - \mu$ and $G m_2 = \mu$ in the dimensionless normalization. For Sun-Earth: $\mu = 3.0404 \times 10^{-6}$. > [!critical] The mass parameter $\mu$ is itself a kinematic measurement > The CR3BP "mass parameter" $\mu = m_2/(m_1 + m_2)$ looks like it requires you to know two independent masses. But this ratio is **never measured by weighing the Sun and the Earth separately**. It is measured by observing how the two bodies orbit their common barycenter. The Sun's reflex motion due to Earth's pull is what determines the ratio. That motion is observed as **positions of the Sun against the background stars over time**, and **positions of the Earth against the background stars over time**. Two orbits, two sets of $(a, T)$. The ratio of the orbital radii about the barycenter equals the inverse mass ratio. Kinematics determines the mass parameter, not the other way around. The number $3.0404 \times 10^{-6}$ comes from observing the geometry of the Earth-Sun-barycenter system, not from a separate dynamical measurement of $m_E$ and $m_S$. See [[mass_ratios]] for the general argument. The CR3BP equations are then propagated in dimensionless form ($\omega = 1$, distance unit = Earth-Sun distance, time unit such that the orbital period is $2\pi$), with a control acceleration $\vec{a}$ applied to keep the spacecraft on the Earth's rotation axis. ### Section 8.2.2 (p.228): The same hovering trick at L1-ish positions The authors design a pole-sitter orbit by specifying the position vector in the rotating frame as a function of time (Eq. 8.5), substituting into Eq. 8.4, and **solving for the controlling acceleration $\vec{a}$ required to make the trajectory work**. This is exactly the hovering procedure from Chapter 3, applied to the three body case. The "dynamics" are inverted: position is the input, acceleration is the output. There is no force balance, only an acceleration prescription. --- ## What the entire book actually computes | Chapter | Topic | What the equation contains | What it does not contain | |---|---|---|---| | 2 | Master equation (Eq. 2.5) | $\ddot{r} = -(\mu/r^3)\vec{r} + \vec{a} + \vec{a}_T$ | mass, force | | 3 | Hovering orbit (Eq. 3.11) | $A = (\mu/r_S^2)(1 - a_S^3/a_T^3)$ | mass, force | | 3 | Modified $\mu$ (Eq. 3.16) | $\mu' = \mu - Ar^2$, "sub-keplerian" | mass, force | | 3 | Force as last step (Eq. 3.21) | $F = m_S A$ | (mass enters here only as scaling) | | 4 | Hill equation (Eq. 4.34-4.39) | positions, mean motion $n$, geometric $b$ | mass, force, $\mu$ | | 4 | E/I vector (Eq. 4.43-4.44) | $\Delta a, l, p, u, s, h$ (lengths and angles) | mass, force, $\mu$ | | 5 | Multi-target rendezvous | "traversing points" (positions in time) | mass, force | | 6 | Initiative approaching | relative position dynamics | mass | | 7 | Fast responsive | flying-around geometry | mass, force | | 8 | CR3BP (Eq. 8.1) | $\ddot{r} + 2\omega\times\dot{r} + \omega\times(\omega\times r) = -\nabla V + \vec{a}$ | mass on the right side, force | | 8 | Mass parameter (Eq. 8.2) | $\mu = m_2/(m_1+m_2)$, dimensionless ratio | independent measurement of $m_1, m_2$ | The book is 248 pages of orbit design, and the spacecraft mass appears in exactly one role across all 248 pages: as a multiplicative scaling at the end of each chapter to convert acceleration into force or fuel. Every equation that determines *where the spacecraft goes* has mass already cancelled. --- ## Application to the null hypothesis **H₁** (assume true): An engineering textbook designed to actually fly spacecraft must use dynamics, with mass and force playing essential roles in the equations. The success of operational space missions is evidence that dynamical variables carry independent physical content. **H₀** (assume false, try to falsify): Even an engineering textbook that designs operational spacecraft missions can be reduced to kinematics, with mass appearing only as a unit conversion at the end. Every equation that propagates a trajectory uses only $\mu$ (a kinematic ratio), positions, velocities, angles, and times. **Result: failed to falsify H₀.** | Test | Claim | What the textbook actually does | Independent dynamics? | |---|---|---|---| | Master equation of motion | "Dynamics" of orbit control | $\ddot{r} = -(\mu/r^3)\vec{r} + \vec{a} + \vec{a}_T$ in pure acceleration units | No | | Hovering orbit thrust | Force needed to hold spacecraft in place | $A = (\mu/r_S^2)(1 - a_S^3/a_T^3)$, mass cancelled | No | | Sub-keplerian orbit | Modified gravitational dynamics | "reduces a part of earth's mass" by changing $\mu$ | No | | Force calculation | Newtonian dynamics in action | $F = m_S A$, last step only, mass is arbitrary input | No | | Hill equation cruising | Linearized two body dynamics | Pure relative kinematics in a rotating frame | No | | E/I vector design | Geometric configuration design | Lengths and angles only | No | | CR3BP propagation | Three body gravitational dynamics | Acceleration equation with kinematic mass ratio | No | | CR3BP mass parameter | Independent body masses | $m_2/(m_1+m_2)$ from barycentric kinematics | No | Eight tests across the entire textbook. Zero independent dynamical variables. **H₀ not falsified.** Zhang, Xu, and Zhou wrote a 248 page astrodynamics textbook in which mass enters the propagation equations zero times, $\mu$ enters as a kinematic ratio measured from satellite tracking, and the design output is always an acceleration profile that gets multiplied by an arbitrarily chosen mass at the end to produce a force budget. They state the dynamics-equals-kinematics conclusion explicitly on page 44 ("reduces a part of earth's mass"), apparently without noticing that this is the entire content of [[First_Principles_Dynamics]] and [[Cosmological_Dynamics_Null]]. --- ## Why this source matters The Bouw 2013 derivation shows that dynamics equals kinematics times $m/m$ from first principles in a centripetal force example. The Artemis II teardown shows that NASA's actual lunar trajectory is computed without independent dynamics. The Textbook Orbital Dynamics note shows that the standard patched conic textbook derivation has no dynamical content. This source extends the argument in two new directions: 1. **Active control case.** Hovering, cruising, and pole-sitting all require continuous thrust. Naively, this should require true dynamics: you have to know the spacecraft mass to compute the force the engine must produce. But the engineering specification is the **acceleration**, not the force. The mass enters only as a downstream scaling for fuel and engine sizing. Active control of spacecraft is a kinematic discipline. 2. **Three body case.** The CR3BP is the textbook example of a "true dynamical" problem where two large bodies create a complicated gravitational landscape and a third small body navigates through it. Even here, the equations are in acceleration form, the "mass parameter" $\mu$ is itself a kinematic ratio measured from observation of barycentric motion, and the design procedure is "specify the trajectory you want, solve for the acceleration required." The CR3BP is a kinematic propagator with a label. This exhausts the obvious places where one might hope to find independent dynamical content in space mission design: closed form trajectories (Artemis, Textbook), active control (Zhang Ch.3), local linearization (Zhang Ch.4), and three body problems (Zhang Ch.8). All four reduce to kinematics. The pattern holds across the spectrum. --- ## Citation Zhang, Yasheng, Xu, Yanli, and Zhou, Haijun (2017). *Theory and Design Methods of Special Space Orbits.* Springer, Singapore. ISBN 978-981-10-2947-9. 248 pages. PDF location: `/home/alan/Downloads/Theory and Design Methods of Special Space Orbits, Yasheng Zhang, Yanli Xu, Haijun Zhou, 2017-1.pdf` --- ## See also - [[Cosmological_Dynamics_Null]] — Parent null hypothesis: GM independence test across 10 bodies - [[First_Principles_Dynamics]] — Bouw's first principles version: dynamics = kinematics times $m/m$ - [[2013_Bouw_Geocentricity_Forces_and_First_Principles]] — Source note on Bouw's mass cancellation argument - [[Textbook_Orbital_Dynamics]] — Patched conic textbook derivation reduced to kinematics - [[Artemis_II_Trajectory]] — NASA's actual Artemis II mission trajectory reduced to kinematics - [[Three_Body_Null]] — For three or more bodies, even the dynamical framework fails - [[mass_ratios]] — Mass as a relational label derived from geometry - [[00_Null_Hypothesis_Index]] — Master null hypothesis index