Abstract. In each of the 10 cases with propagators of unit or zero mass, the finite part of the scalar 3-loop tetrahedral vacuum diagram is reduced to 4-letter words in the 7-letter alphabet of the 1-forms $\Omega:=dz/z$ and $\omega_p:=dz/ (\lambda^{-p}-z)$, where $\lambda$ is the sixth root of unity. Three diagrams yield only $\zeta(\Omega^3\omega_0)=\frac1{90}\pi^4$. In two cases $\pi^4$ combines with the Euler-Zagier sum $\zeta(\Omega^2\omega_3\omega_0)=\sum_{m> n>0}(-1)^{m+n}/m^3n$; in three cases it combines with the square of Clausen's ${\rm Cl}_2(\pi/3)=\Im\,\zeta(\Omega\omega_1)=\sum_{n>0}\sin(\pi n/3)/n^2$. The case with 6 masses involves no further constant; with 5 masses a Deligne-Euler-Zagier sum appears: ${\frak R} \zeta(\Omega^2\omega_3\omega_1)= \sum_{m>n>0}(-1)^m\cos(2\pi n/3)/m^3n$. The previously unidentified term in the 3-loop rho-parameter of the standard model is merely $D_3=6\zeta(3)-6{\rm Cl}_2^2(\pi/3)-\frac{1}{24}\pi^4$. The remarkable simplicity of these results stems from two shuffle algebras: one for nested sums; the other for iterated integrals. Each diagram evaluates to 10 000 digits in seconds, because the primitive words are transformable to exponentially convergent single sums, as recently shown for $\zeta(3)$ and $\zeta(5)$, familiar in QCD. Those are SC$^*(2)$ constants, whose base of super-fast computation is 2. Mass involves the novel base-3 set SC$^*(3)$. All 10 diagrams reduce to SC$^*(3)\cup$SC$^* (2)$ constants and their products. Only the 6-mass case entails both bases.
[1]
Nicolai Reshetikhin,et al.
Quantum Groups
,
1993
.
[2]
G. G. Stokes.
"J."
,
1890,
The New Yale Book of Quotations.
[3]
F. M. Saxelby.
Experimental Mathematics
,
1902,
Nature.
[4]
Leonard Lewin,et al.
Polylogarithms and Associated Functions
,
1981
.
[5]
T. Hall,et al.
Journal of Algebra
,
1964,
Nature.
[6]
C. Caldwell.
Mathematics of Computation
,
1999
.
[7]
Jonathan M. Borwein,et al.
Special Values of Multidimensional Polylogarithms
,
2001
.
[8]
David J. Broadhurst,et al.
Polylogarithmic ladders, hypergeometric series and the ten millionth digits of ?(3) and ?(5)
,
1998
.