Using pulsar timing arrays and the quantum normalization condition to constrain relic gravitational waves

In the non-standard model of relic gravitational waves (RGWs) generated in the early universe, the theoretical spectrum of is mainly described by an amplitude $r$ and a spectral index $\beta$, the latter usually being determined by the slope of the inflation potential. Pulsar timing arrays (PTAs) data have imposed constraints on the amplitude of strain spectrum for a power-law form as a phenomenological model. Applying these constraints to a generic, theoretical spectrum with $r$ and $\beta$ as independent parameters, we convert the PTAs constraint into an upper bound on the index $\beta$, which turns out to be less stringent than those upper bounds from BBN, CMB, and LIGO/VIRGO, respectively. Moreover, it is found that PTAs constrain the non-standard RGWs more stringent than the standard RGWs. If the condition of the quantum normalization is imposed upon a theoretical spectrum of RGWs, $r$ and $\beta$ become related. With this condition, a minimum requirement of the horizon size during inflation is greater than the Planck length results in an upper bound on $\beta$, which is comparable in magnitude to that by PTAs. When both PTAs and the quantum normalization are applied to a theoretical spectrum of RGWs, constraints can be obtained for other cosmic processes of the early universe, such as the reheating, a process less understood observationally so far. The resulting constraint is consistent with the slow-roll, massive scalar inflation model. The future SKA will be able to constrain RGWs further and might even detect RGWs, rendering an important probe to the very early universe.