Cosmic string in Abelian-Higgs model with enhanced symmetry - Implication to the axion domain-wall problem

In our previous work, we found new types of the cosmic string solutions in the Abelian-Higgs model with an enhanced U(1) global symmetry. We dubbed those solutions as the compensated/uncompensated strings. The compensated string is similar to the conventional cosmic string in the Abrikosov-Nielsen-Olesen (ANO) string, around which only the would-be Nambu-Goldstone (NG) boson winds. Around the uncompensated string, on the other hand, the physical NG boson also winds, where the physical NG boson is associated with the spontaneous breaking of the enhanced symmetry. Our previous simulation in the 2+1 dimensional spacetime confirmed that both the compensated/uncompensated strings are formed at the phase transition of the symmetry breaking. Non-trivial winding of the physical NG boson around the strings potentially causes the so-called axion domain- wall problem when the model is applied to the axion model. In this paper, we perform simulation in the 3+1 dimensional spacetime to discuss the fate of the uncompensated strings. We observe that the evolution of the string-network is highly complicated in the 3+1 dimensional simulation compared with that seen in the previous simulation. Despite such complications, we find that the number of the uncompensated strings which could cause can be highly suppressed at late times. Our observation suggests that the present setup can be applied to the axion model without suffering from the axion domain-wall problem.

General purpose graphics-processing-unit implementation of cosmological domain wall network evolution

Topological defects unavoidably form at symmetry breaking phase transitions in the early universe. To probe the parameter space of theoretical models and set tighter experimental constraints (exploiting the recent advances in astrophysical observations), one requires more and more demanding simulations, and therefore more hardware resources and computation time. Improving the speed and efficiency of existing codes is essential. Here we present a general purpose graphics-processing-unit implementation of the canonical Press-Ryden-Spergel algorithm for the evolution of cosmological domain wall networks. This is ported to the Open Computing Language standard, and as a consequence significant speedups are achieved both in two-dimensional (2D) and 3D simulations.