78518dd2e522193a.tex
1: \begin{abstract}
2: \centerline{\bf Abstract \vspace{0.3cm}}
3: Perturbation expansions appear to
4: be divergent series  in many physically interesting situations, including in quantum field theories like quantum electrodynamics (QED) and quantum chromodynamics  (QCD), where the perturbative coefficients exhibit a factorial growth at large orders.   While this  feature 
5: has little impact on physical predictions in QED,  it can have nontrivial consequences in applications of perturbative QCD at moderate energies. In particular, it affects the   theoretical error in the extraction of the strong coupling $\alpha_s$ from hadronic  $\tau$  decays, despite progress of perturbative calculations available at present to four loops.  We discuss a new  type of perturbative expansion for QCD correlators, which uses instead of the standard powers of the  coupling  a new set of expansion functions. These functions are defined by means of an
6: optimal conformal mapping of the Borel complex plane, which implements the known features of the high-order divergence in terms of the lowest Borel-plane singularities. 
7: The  properties of the expansion functions resemble those of the expanded correlators, by exhibiting in particular the singular behaviour of the correlators at $\alpha_s=0$.  We prove the good convergence properties of the new expansions on mathematical models that simulate  the physical polarization function for light quarks  and its derivative (the Adler function), in various prescriptions of renormalization-group summation. 
8: 
9: \end{abstract}
10: