Abstract:This work derives 5 methods to evaluate families of odd zeta values by combining a power of $\pi$ with Lambert series whose ratios of successive terms tend to $e^{-\pi\sqrt{a}}$ with integers $a\ge7$, outperforming Ramanujan's results with merit $a=4$. Families with $a=7$ and $a=8$ evaluate $\zeta(2n+1)$. Families with $a=9$ and $a=16$ evaluate $\zeta(4n+1)$ with faster convergence. A fifth family with $a=12$ evaluates $\zeta(6n+1)$ and gives the fastest convergence for $\zeta(6n+7)$. Members of three of the families were discovered empirically by Simon Plouffe. An intensive new search strongly suggests that there are no more than 5 families with integers $a\ge7$. There are at least 20 families that involve Lambert series with rational $a>4$. Quasi-modular transformations of Lambert series resolve rational sequences that were discovered empirically. Expansions of Lambert series in polylogarithms, familiar from quantum field theory, provide proofs of all known evaluations with rational merit $a>4$.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to propose and verify five new methods for rapidly converging the calculation of odd - numbered zeta values (the values of the Riemann zeta function at odd points), and these methods are more efficient than Ramanujan's previous results. Specifically, by combining powers of π and Lambert series (the ratio of whose consecutive terms tends to \(e^{-\pi\sqrt{a}}\), where the integer \(a\geq7\)), the author has derived five types of rapidly converging evaluation methods for odd - numbered zeta values.
### Main problem description:
1. **Faster convergence speed**: The methods proposed in the paper can converge to odd - numbered zeta values faster than Ramanujan's results. For example, for \( \zeta(3)\) and \( \zeta(5)\), Ramanujan's results have the advantage of \(a = 4\), while the new methods reach the advantage of \(a\geq7\).
2. **Discovery of five families**: The author has discovered five different families, each corresponding to a specific value of \(a\) (7, 8, 9, 16 and 12 respectively), and these families can be used to evaluate different types of odd - numbered zeta values:
- The families with \(a = 7\) and \(a = 8\) are used to evaluate \( \zeta(2n + 1)\).
- The families with \(a = 9\) and \(a = 16\) are used to evaluate \( \zeta(4n+1)\), and the convergence speed is faster.
- The family with \(a = 12\) is used to evaluate \( \zeta(6n + 1)\), and provides the fastest convergence speed for \( \zeta(6n+7)\).
3. **Mathematical proof and verification**: The paper not only proposes these methods, but also shows the effectiveness of these methods through strict mathematical proofs, and further confirms the existence and uniqueness of these five families through computer - aided search.
### Markdown representation of formulas:
- The general form of the Lambert series is:
\[
S_k(q)=\sum_{n = 1}^{\infty}\frac{1}{n^k(q^n - q^{-n})}=\sum_{m = 1}^{\infty}\text{Li}_k(q^m)
\]
where \( \text{Li}_k(x)=\sum_{n>0}\frac{x^n}{n^k}\) is the polylogarithm function.
- For the family with \(a = 7\), we have:
\[
\zeta(k)=A_{k,0}\sqrt{7}\pi^k+\sum_{d\in\{1,2,4\}}A_{k,d}S_k(q_d^7),\quad q_7=\exp(-\pi\sqrt{7})
\]
- For the family with \(a = 8\), we have:
\[
\zeta(k)=B_{k,0}\sqrt{2}\pi^k+\sum_{d\in\{2,3,6\}}B_{k,d}S_k(q_d^2),\quad q_2=\exp(-\pi\sqrt{2})
\]
- For the family with \(a = 9\), we have:
\[
\zeta(k)=C_{k,0}\pi^k+\sum_{d\in\{3,4,6,12\}}C_{k,d}S_k(q_d^1),\quad q_1=\exp(-\pi)
\]
- For the family with \(a = 16\), we have:
\[
\zeta(k)=D_{k,0}\pi^k+\sum_{d\in\{4,5,10,20\}}D_{k,d}S_k(q_d^1),\quad q_1=\exp(-\pi)