Cosmic acceleration from asymmetric branes

TL;DR

The paper investigates cosmic acceleration without a brane cosmological constant by placing a single 3-brane between two distinct Gauss-Bonnet bulks, introducing weighted junction conditions that depend on the asymmetry. It analyzes two explicit models—the generalized Randall-Sundrum setup and a stringy Gauss-Bonnet scenario—focusing on symmetric and antisymmetric branches. It demonstrates that late-time de Sitter expansion can emerge in antisymmetric configurations without vacuum energy, and that, in the RS-like case, standard cosmology can be recovered back to nucleosynthesis with appropriate parameter relations. The stringy model yields acceleration under specific inequalities, while the (-) branch can fail to reproduce standard early-time cosmology despite exhibiting acceleration. Collectively, the work shows that IR modifications of gravity via asymmetric branes can drive cosmic acceleration geometrically, with implications for gravity localization and holography, albeit with tuning and singularity considerations.

Abstract

We consider a single 3-brane sitting in between two different five dimensional spacetimes. On each side of the brane, the bulk is a solution to Gauss-Bonnet gravity, although the bare cosmological constant, funda mental Planck scale, and Gauss-Bonnet coupling can differ. This asymmetry leads to weighted junction conditions across the brane and interesting brane cosmology. We focus on two special cases: a generalized Randall-Sundrum model without any Gauss-Bonnet terms, and a stringy model, without any bare cosmological constants, and positive Gauss-Bonnet coupling. Even though we assume there is no vacuum energy on the brane, we find late time de Sitter cosmologies can occur. Remarkably, in certain parameter regions, this acceleration is preceded by a period of matter/radiation domination, with $H^2 \propto ρ$, all the way back to nucleosynthesis.