Polyhedral Potential and Variational Integrator Computation of the Full Two Body Problem

Eugene G. Fahnestock, Taeyoung Lee, Melvin Leok, N. Harris McClamroch, Daniel J. Scheeres

TL;DR

This paper introduces a robust framework for simulating the "Full Two Body Problem"—the complex gravitational interaction between two irregular rigid bodies, such as binary asteroids. By combining a polyhedral gravity model with a Lie Group Variational Integrator (LGVI), the authors created a simulation that is both computationally efficient and physically precise, preserving essential geometric properties like total energy and angular momentum over long timescales.

What Problem Does This Paper Address?

In classical orbital mechanics, we often treat planets and moons as point masses or perfect spheres. However, for objects like binary asteroids or a large spacecraft orbiting a small moon, shape matters. These bodies are "lumpy," and their irregular mass distributions cause their rotational motion (attitude) and translational motion (orbit) to become coupled. A change in how an asteroid spins can actually alter its orbital path.

Simulating this "Full Two Body Problem" (F2BP) is notoriously difficult for two reasons:

  1. Potential Complexity: Standard mathematical tools like spherical harmonics often fail to converge when two irregular bodies get very close to each other.
  2. Numerical Drift: Common integration methods (like Runge-Kutta) do not respect the underlying geometry of the rotation group. Over time, they "leak" energy or cause rotation matrices to lose their orthogonality, leading to unphysical results in long-term simulations.

Methods

The authors tackle these challenges by synthesizing two sophisticated tools:

  • Polyhedral Potential Modeling: Instead of point masses, the bodies are represented as polyhedra composed of "simplices" (tetrahedra). The mutual gravitational potential is calculated using a Legendre-series expansion. This method is flexible, allowing for varying densities and complex shapes (craters, clefts), and it remains valid even when the bodies are in close proximity.
  • Lie Group Variational Integrator (LGVI): This is the core innovation for the dynamics. Rather than discretizing the continuous equations of motion, the LGVI discretizes Hamilton’s principle (the principle of least action) directly. Because it operates on the Lie group $SE(3)$ (the configuration space for rigid body motion), it naturally preserves the structure of rotation matrices without needing "hacks" like re-normalization. It is a symplectic integrator, meaning it maintains the qualitative behavior of the system (like energy stability) for exponentially long periods.

Datasets

As this is a theoretical and computational mechanics paper, it does not use external "training" datasets. Instead, the authors evaluate their method using:

  • Synthetic Octahedral Models: Two octahedral rigid bodies (8 faces each) were used as test cases. These are the simplest polyhedral shapes that can demonstrate coupled orbit-attitude dynamics.
  • Four Simulation Scenarios: The authors defined specific initial conditions (position, velocity, spin) to test different behaviors, including medium-eccentricity orbits, close encounters leading to orbital disruption, and highly unstable long-duration trajectories.

Key Insights and Ideas

  • Geometric Integrity: The most significant "aha moment" is that preserving the structure of the problem is more important than raw high-order accuracy. The LGVI respects the $SE(3)$ group structure, ensuring that "rotation matrices stay rotation matrices" throughout the simulation.
  • Efficiency through Simplicity: While high-order methods like Runge-Kutta-Fehlberg 7(8) require 14 force/torque evaluations per time step, the LGVI achieves comparable or better long-term results with only one evaluation per step. Since gravity calculations are the most expensive part of the code, this leads to massive speedups.
  • Coupling is Non-Negligible: The simulations show that for irregular bodies, the orbit and attitude are inseparable. In one scenario, the mutual interaction during a close encounter provides enough energy transfer to kick the bodies from an elliptical orbit into a hyperbolic escape trajectory—a phenomenon that point-mass models completely miss.

Key Results

  • Long-term Stability: In a long-duration test ($5 \times 10^6$ seconds), the LGVI maintained total energy errors at a level of $10^{-11}$ with no mean drift. In contrast, the standard RKF7(8) method showed linear energy drift and became "completely unreliable" as its rotation matrix errors grew to unacceptable levels.
  • Computational Performance: The LGVI proved to be significantly more efficient. For a given error tolerance, the LGVI could use larger time steps and fewer force evaluations than traditional methods.
  • Parallelization: The authors demonstrated that the polyhedral potential calculations are "embarrassingly parallel," achieving two orders of magnitude reduction in computation time using MPI on a cluster.

Limitations and Future Work

  • Shape Complexity: The paper primarily uses simple octahedra. While the math scales to more complex meshes (thousands of faces), the computational cost of the double-summation in the potential calculation would increase significantly.
  • Rigidity Assumption: The model assumes perfectly rigid bodies. The authors briefly mention "rubble-pile" models, suggesting that future work could extend this framework to fragmented or deformable bodies.
  • Legendre Convergence: While better than spherical harmonics, the Legendre series still requires more terms as bodies get closer, which could become a bottleneck for contact or near-contact scenarios.

Summary updated

Abstract

We present a combination of tools which allows for investigation of the coupled orbital and rotational dynamics of two rigid bodies with nearly arbitrary shape and mass distribution, under the influence of their mutual gravitational potential. Methods for calculating that mutual potential and resulting forces and moments for a polyhedral body representation are simple and ecient. Discrete equations of motion, referred to as the Lie Group Variational Integrator (LGVI), preserve the structure of the configuration space, SE(3), as well as the geometric features represented by the total energy and the total angular momentum. The synthesis of these approaches allows us to simulate the full two body problem accurately and eciently. Simulation results are given for two octahedral rigid bodies for comparison with other integration methods and to show the qualities of the results thus obtained. A significant improvement is seen over other integration methods while correctly capturing the interesting eects of strong orbit and attitude dynamics coupling, in multiple scenarios.
BibTeX
@inproceedings{fahnestock2006polyhedral,
  author = {Eugene G. Fahnestock and Taeyoung Lee and Melvin Leok and N. Harris McClamroch and Daniel J. Scheeres},
  title = {{Polyhedral Potential and Variational Integrator Computation of the Full Two Body Problem}},
  booktitle = {AIAA/AAS},
  year = {2006},
  doi = {10.2514/6.2006-6289},
}

Loading related content...

Figure