On Synthetic Gravitational Waves from Multi-field Inflation

TL;DR

This paper investigates the potential for observable gravitational waves generated during multi-field inflation through particle production, showing that detectable signals are still feasible within current observational constraints.

Contribution

It provides a detailed calculation of scalar correlators in a multi-field inflation model, demonstrating that large tensor modes are compatible with existing CMB limits.

Findings

Possible to achieve tensor-to-scalar ratio r ≈ 10^{-3}

Current CMB data constraints are compatible with significant gravitational wave signals

Revised understanding of limitations in particle production models during inflation

Abstract

We revisit the possibility of producing observable tensor modes through a continuous particle production process during inflation. Particularly, we focus on the multi-field realization of inflation where a spectator pseudo-scalar $\sigma$ induces a significant amplification of the ${\rm U}(1)$ gauge fields through the coupling $\propto \sigma F_{\mu\nu}\tilde{F}^{\mu\nu}$. In this model, both the scalar $\sigma$ and the Abelian gauge fields are gravitationally coupled to the inflaton sector, therefore they can only affect the primordial scalar and tensor fluctuations through their mixing with gravitational fluctuations. Recent studies on this scenario show that the sourced contributions to the scalar correlators can be dangerously large to invalidate a large tensor power spectrum through the particle production mechanism. In this paper, we re-examine these recent claims by explicitly calculating the dominant contribution to the scalar power and bispectrum. Particularly, we show that once the current limits from CMB data are taken into account, it is still possible to generate a signal as large as $r \approx 10^{-3}$ and the limitations on the model building are more relaxed than what was considered before.