A Review of Methods for Solving the Many-Body Problem: From the Two-Body Analytical Solution to Four Alternative Approaches

Author: Zhang Suhang, Luoyang, Henan

Abstract

The two-body problem has a global closed-form analytical solution, but this special case has long been treated as the standard for all dynamical systems. When three-body and many-body systems cannot yield a similar solution in terms of finite elementary functions, the saying that "the many-body problem is unsolvable" has emerged. This paper reviews the origin of this saying and points out that it stems from a narrow understanding of the word "solvable." It then summarizes four practical approaches by which humanity handles the many-body problem: numerical solution, perturbation approximation, Fourier series approximation, and statistical mean-field methods. These four approaches each have their own scope of applicability and jointly cover various scenarios from few-body to many-body, from short-time to long-time, and from individual to collective behavior. Many-body systems continue to evolve physically; the so-called "unsolvability" merely means that no global closed-form solution in Newton's sense exists, and does not imply that the system itself has no determinate behavior.

Keywords: three-body problem; many-body dynamics; analytical solution; numerical solution; perturbation; statistical mechanics

1 Introduction

Newton obtained a perfect analytical solution for the two-body problem: when two point masses are acted upon only by gravity, their trajectories can be written out completely in terms of a finite number of elementary functions. This is one of the most successful examples in classical mechanics.

However, when this method was extended to three or more celestial bodies, it was found that a similar finite-elementary global closed-form solution could no longer be constructed. In the late 19th century, Poincaré proved that no such analytical solution exists for the three-body system. Thereafter, the sayings "the three-body problem is unsolvable" and "the many-body problem has no solution" gradually became popular.

The question is: what exactly does "unsolvable" mean here? Does it mean that the system itself has no determinate law of evolution, or merely that a certain specific mode of mathematical expression is no longer applicable?

This paper argues that the latter understanding is more consistent with reality. Natural systems are always evolving; celestial systems continue to operate; there is no "unsolvable" universe. The so-called "unsolvability" is merely the failure of the Newtonian analytical standard in the many-body case, not a defect in the laws of nature themselves. The following sections will review the origin of this standard and summarize the four methods humanity actually uses to solve many-body problems.

2 The Origin of the Saying "The Many-Body Problem Is Unsolvable"

2.1 Newton's Two-Body Paradigm

The two-body analytical solution given by Newton in The Mathematical Principles of Natural Philosophy has several notable features: a finite number of elementary functions, global validity, no series, and no approximation. This solution is mathematically concise and complete, and it left a deep impression.

Newton himself had a strong expectation of this, believing that any determinate natural system should possess such a solution. In his later years he devoted great effort to studying the three-body problem, attempting to find a similar closed-form solution, but without success.

From today's perspective, the fact that the two-body problem admits a global closed-form solution is a low-dimensional special case. Between two point masses there is only one pair of interactions, with no cross-coupling, and the equations of motion are integrable. The situation is different for three or more bodies: all coupling terms are switched on, the system is no longer integrable, and the two-body method cannot be directly applied.

2.2 Poincaré's Conclusion and Its Boundaries

In studying the three-body problem, Poincaré proved that the three-body system has no global closed-form solution composed of a finite number of elementary functions, that the system is not completely integrable, and that chaotic behavior highly sensitive to initial conditions exists.

This conclusion itself is rigorous. But it has clear boundaries: it negates only the existence of a "finite-elementary global closed-form solution," and does not negate that the system has a determinate evolutionary trajectory, nor does it negate that the system's behavior can be described and calculated in other ways.

When later generations cite Poincaré's conclusion, they often cross these boundaries and directly equate "no Newtonian analytical solution" with "no solution." This simplification is inaccurate.

2.3 A Confusion That Needs to Be Clarified

The saying "no solution" actually conflates two different levels of the problem:

1. The physical level: given initial conditions, does the system have a unique, continuous evolution?

2. The symbolic level: can humans write out this evolution completely in terms of a finite number of elementary functions?

The answer at the physical level is yes: many-body systems have determinate evolution, and the galaxies, star clusters, and planetary systems in the universe are all continuously operating, which itself is proof that the system has a solution.

The answer at the symbolic level is no: no global closed-form solution in Newton's sense exists.

Presenting the symbolic-level negation as a physical-level "no solution" is the main reason why the saying "the many-body problem is unsolvable" has long been popular.

3 Four Practical Approaches by Which Humanity Handles the Many-Body Problem

Since no Newtonian analytical solution exists, how does humanity actually handle the many-body problem? In summary, there are mainly four approaches, each with its own scope of applicability.

3.1 Numerical Solution

Abandon the global closed-form solution and instead compute step by step along time.

The method is: given initial conditions, use numerical methods to advance time step by step and obtain the trajectory within any finite time. The step size can be adjusted, the precision can be controlled, and the error can be estimated.

This path does not give a unified formula, but it can give the specific trajectory of a specific problem. Spacecraft orbit calculations, predictions of celestial positions, and many-body simulations in engineering basically all rely on this method.

Its essence is to replace the Newtonian global continuous formula with discrete computation in time.

3.2 Perturbation Approximation

For weakly coupled many-body systems—for example, in the solar system the interactions among planets are tiny relative to the Sun's gravity—correction terms can be superimposed on the two-body solution.

The method is: first take the two-body analytical solution as the basis, then treat the gravity of other celestial bodies as a small quantity and add corrections order by order. The higher the order, the closer the result is to the true orbit.

This path does not seek a one-step closed-form solution, but handles many-body coupling by successive approximation. Its scope of applicability is systems with relatively weak coupling and clear hierarchy.

3.3 Fourier Series Approximation

For quasi-periodic many-body orbits, the motion can be decomposed into a superposition of multiple periodic motions.

The method is: expand the orbit into a Fourier series and truncate the number of terms according to the required precision. The more terms, the more accurate the approximation.

The premise of this path is that the orbit is quasi-periodic. For chaotic orbits, this method is not applicable. Its core idea is: instead of requiring a finite closed-form expression, use a finite truncation of an infinite series to approximate the true solution.

3.4 Statistical Methods and Mean Field

When the number of particles or celestial bodies is extremely large, solving trajectories one by one is both impossible and unnecessary. At this point one turns to describing the collective macroscopic laws.

The method is: do not track the coordinates of individual particles, but instead solve for statistical quantities such as density distribution, velocity distribution, and mean field. The more particles there are, the more stable the statistical laws.

Galactic dynamics, plasma physics, and research on the large-scale structure of the universe mainly rely on this method. It is the practical scheme for handling massive many-body systems.

4 Scope of Applicability of the Four Approaches

The above four approaches are not substitutes for one another, but each covers a different scenario:

Scenario | Method

Few-body, strongly coupled, short-time behavior | Numerical solution

Weakly coupled, clearly hierarchical celestial systems | Perturbation approximation

Quasi-periodic regular orbits | Fourier series approximation

Collective behavior of massive numbers of particles or celestial bodies | Statistical methods and mean field

Newton's two-body analytical solution covers only one special case among them. Treating this special case as the standard for all dynamical systems is the root of the saying "the many-body problem is unsolvable."

5 Conclusion

This paper reviews the origin of the saying "the many-body problem is unsolvable" and summarizes the four methods humanity actually uses to solve many-body problems. The core conclusions are as follows:

1. The global closed-form solution of the two-body problem is a low-dimensional special case and does not possess natural generalizability to many-body systems.

2. What Poincaré proved is that the three-body system has no finite-elementary global closed-form solution; he did not negate the determinate evolution of the system itself.

3. The saying "the many-body problem is unsolvable" conflates solvability at the physical level with solvability at the symbolic level.

4. Humanity actually uses four approaches—numerical solution, perturbation approximation, Fourier series approximation, and statistical methods and mean field—covering various scenarios from few-body to many-body, from short-time to long-time, and from individual to collective behavior.

5. The so-called "unsolvability" merely means that no analytical solution in Newton's sense exists, and does not imply that many-body systems have no determinate behavior.

Natural systems exist independently of human formulas, and the ways in which humanity handles the many-body problem need not be limited to the single standard of the two-body analytical solution.

References

(Omitted)