# Is it true that the 3 body problem can't be solved using the four basic functions, radicals, and integrals?

The two-body problem can be completely solved via two one-body problems, which only uses the four basic binary functions. However, the three-body problem cannot be solved with these functions and first-order integrals. So I am wondering, is there any finite numerical solution to the three-body problem?

• Israel's Technion researchers have found an effective solution to the famous age-old, three-body problem in physics. in a paper recently published in physical review with the current study by ginat and perets, the entire, multi-stage, three-body interaction is fully solved, statistically. scitechdaily.com/a-centuries-old-physics-mystery-solved
– Alex
Dec 10, 2021 at 2:04
• I just ran across this and the answer. I can't up vote either of you again, but I wonder if the answer is good enough to be accepted?
– uhoh
Feb 3 at 10:15
• @uhoh Yes, just waited a little too long to accept the answer and it got lost in my inbox. Feb 3 at 13:38
• @fasterthanlight I do that too, and in fact let many of my questions "age" for months or even longer sometimes. This one is so cool however that I'm really loving it! I added a comment under DH's answer btw.
– uhoh
Feb 3 at 13:50

In a sense, even solving the two body problem as a function of time is unsolvable in terms of the elementary functions. The problem is that the solution involves the solving for the inverse of Kepler's problem, $$M = E - e \sin E$$. This inverse function is transcendental and cannot be expressed in terms of the elementary functions. That said, it's fairly easy to solve for $$E$$ to an arbitrary degree of precision.
Those restricted three body problem Lagrange points are special cases. The generic case of the three body problem is notoriously unsolvable in terms of the elementary functions. There is an infinite series solution, but nobody uses it. Over a century ago there was a prize for solving the three body problem. Karl Frithiof Sundman was awarded that prize in 1912 for showing that a solution exists in the form of an infinite series in which the terms are ever increasing powers of the cube root of time. The reason no one uses this solution is that the number of required terms can be huge. Eight million terms is not near enough. It's more on the order of $$10^{8000000}$$ terms.