Renormalized Finite Temperature phi^4 theory from the 2PI Effective Action
We present an analytical and numerical study of scalar phi^4 theory at finite temperature with a renormalized 2-loop truncation of the 2PI effective action.
arXiv subjects
Publications and source records attributed to A. Arrizabalaga.
We present an analytical and numerical study of scalar phi^4 theory at finite temperature with a renormalized 2-loop truncation of the 2PI effective action.
We discuss the method of $Φ$-derivable approximations in gauge theories. There, two complications arise, namely the violation of Bose symmetry in correlation functions and the gauge dependence. For the latter we argue that the error introduced by the gauge dependent terms is controlled, therefore not invalidating the method.
We examine the problem of gauge dependence of the 2PI effective action and its Phi-derivable approximations in gauge theories. The dependence on the gauge-fixing condition is obtained. The result shows that Phi-derivable approximations, defined as truncations of the 2PI effective action at a certain order, have a controlled gauge dependence, i.e. the gauge dependent terms appear at higher order than the truncation order. Furthermore, using the stationary point obtained for the approximation to evaluate the complete 2PI effective action boosts the order at which the gauge dependent terms appear to twice the order of truncation. We also comment on the significance of this controlled gauge dependence.
We show that the logarithmic divergences that appear in the classical approximation of the finite temperature SU(N) self-energy are transverse. We use the Ward identities in linear gauges and the fact that the superficial degree of divergence d of a classical diagram only depends on the number of loops l via d=2-l. We comment on the relevance of this result to the construction of a low-energy effective theory beyond HTLs.