The phenomenology of trapped inflation

TL;DR

This paper refines the theoretical understanding of trapped inflation by updating equations, constraints, and numerical results, revealing its viability with certain potentials and specific non-Gaussianity signatures.

Contribution

It extends the equations for trapped inflation perturbations to second order, updates constraints, and provides more accurate predictions for the power spectrum and bispectrum.

Findings

Monomial potentials can fit observational data.

Inflaton can remain sub-Planckian during observable inflation.

Gravitational waves are too small to detect soon.

Abstract

Trapped inflation is a mechanism in which particle production from the moving inflaton is the main source of friction in the inflaton equation of motion. The produced fields source inflaton perturbations, which dominate over the vacuum ones. We employ the set of equations for the inflaton zero mode and its perturbations which was developed in the original work on trapped inflation, and which we extend to second order in the perturbations. We build on this study by updating the experimental constraints, and by replacing the existing approximate solutions with more accurate ones. We obtain a different numerical value for the amplitude of the scalar power spectrum, and a parametrically different result for the bispectrum. This has implications for the allowed region of parameter space in models of trapped inflation, and for some of the phenomenological results obtained in this region. The main results in the allowed region are the following: monomial inflaton potentials, such as $V \propto \varphi ,\, \varphi^2$ can be compatible with the data, and (in a portion of the allowed region) the inflaton can be sub-Planckian over all the "observable" stage of inflation; gravitational waves from this mechanism are too small to be observed in the foreseeable future; a non-gaussianity parameter $-60 < f_{\rm NL} < -20 $ is obtained in the allowed region for exactly equilateral configurations, with a mild dependence on the parameters of the model and of the inflaton potential.