Cosmological perturbations across a curvature bounce

TL;DR

This paper investigates the evolution of cosmological inhomogeneities through a non-singular curvature bounce inspired by string theory, comparing numerical results with analytic predictions and emphasizing the importance of matching conditions for the curvature perturbation.

Contribution

It introduces a string-inspired, non-local dilaton potential model for a non-singular bounce and analyzes inhomogeneity evolution with a focus on matching conditions for curvature perturbations.

Findings

Good agreement with analytic expressions when matching for al{R}

Continuity of al{R} across the bounce is crucial

Bardeen potential matching yields less accurate results

Abstract

String-inspired cosmologies, whereby a non-singular curvature bounce is induced by a general-covariant, $T$-duality-invariant, non-local dilaton potential, are used to study numerically how inhomogeneities evolve and to compare the outcome with analytic expressions obtained through different matching conditions across the bounce. Good agreement is found if continuity across the bounce is assumed to hold for $\cal{R}$, the curvature perturbation on comoving hypersurfaces, rather than for the Bardeen potential.