We study cosmological perturbation theory within the framework of unimodular gravity. We show that the Lagrangian constraint on the determinant of the metric required by unimodular gravity leads to an extra constraint on the gauge freedom of the metric perturbations. Although the main equation of motion for the gravitational potential remains the same, the shift variable, which is gauge artifact in General Relativity, cannot be set to zero in unimodular gravity. This non-vanishing shift variable affects the propagation of photons throughout the cosmological evolution and therefore modifies the Sachs-Wolfe relation between the relativistic gravitational potential and the microwave temperature anisotropies. However, for adiabatic fluctuations the difference between the result in General Relativity and unimodular gravity is suppressed on large angular scales. Thus, no strong constraints on the theory can be derived.
We study the preheating process in a model of DBI inflation with a DBI-type inflaton coupling to a canonical entropy field. At the end of inflation, the inflaton field oscillates around its vacuum which can arise from an infrared cutoff parameter on the warp factor and correspondingly the evolution of its fluctuations can be approximately described by a generalized Hills equation in third order. We study the field fluctuations numerically and show that they could grow exponentially since the instability bands commonly exist in the DBI models if the amplitudes of background oscillations are of order or larger than the cutoff parameter. Our numerical result also reveals that the particle excitation of the matter field is more dramatic than that in usual case since the parametric resonance lasts longer when the effect of a warp factor is taken into account. Therefore, we conclude that the preheating process in the model of DBI inflation could be more efficient than that in standard inflation models.
We study cosmological evolutions of the generalized model of nonlinear massive gravity in which the graviton mass is given by a rolling scalar field and is varying along time. By performing dynamical analysis, we derive the critical points of this system and study their stabilities. These critical points can be classified into two categories depending on whether they are identical with the traditional ones obtained in General Relativity. We discuss the cosmological implication of relevant critical points.
In this paper we revisit the dynamical dark energy model building based on single scalar field involving higher derivative terms. By imposing a degenerate condition on the higher derivatives in curved spacetime, one can select the models which are free from the ghost mode and the equation of state is able to cross the cosmological constant boundary smoothly, dynamically violate the null energy condition. Generally the Lagrangian of this type of dark energy models depends on the second derivatives linearly. It behaves like an imperfect fluid, thus its cosmological perturbation theory needs to be generalized. We also study such a model with explicit form of degenerate Lagrangian and show that its equation of state may cross -1 without any instability.
In this paper we study the evolution of cosmological perturbations in the presence of dynamical dark energy, and revisit the issue of dark energy perturbations. For a generally parameterized equation of state (EoS) such as w_D(z) = w_0+w_1frac{z}{1+z}, (for a single fluid or a single scalar field ) the dark energy perturbation diverges when its EoS crosses the cosmological constant boundary w_D=-1. In this paper we present a method of treating the dark energy perturbations during the crossing of the $w_D=-1$ surface by imposing matching conditions which require the induced 3-metric on the hypersurface of w_D=-1 and its extrinsic curvature to be continuous. These matching conditions have been used widely in the literature to study perturbations in various models of early universe physics, such as Inflation, the Pre-Big-Bang and Ekpyrotic scenarios, and bouncing cosmologies. In all of these cases the EoS undergoes a sudden change. Through a detailed analysis of the matching conditions, we show that delta_D and theta_D are continuous on the matching hypersurface. This justifies the method used[1-4] in the numerical calculation and data fitting for the determination of cosmological parameters. We discuss the conditions under which our analysis is applicable.
