We discuss the appearance at the QCD phase transition, and the subsequent decay, of axion walls bounded by strings in N = 1 axion models. We argue on intuitive grounds that the main decay mechanism is into barely relativistic axions. We present numerical simulations of the decay process. In these simulations, the decay happens immediately, in a time scale of order the light travel time, and the average energy of the radiated axions is ω a ≃ 7m a for v a /m a ≃ 500. ω a is found to increase approximately linearly with ln(v a /m a ). Extrapolation of this behaviour yields ω a ∼ 60 m a in axion models of interest. We find that the contribution to the cosmological energy density of axions from wall decay is of the same order of magnitude as that from vacuum realignment, with however large uncertainties. The velocity dispersion of axions from wall decay is found to be larger, by a factor 10 3 or so, than that of axions from vacuum realignment and string decay. We discuss the implications of this for the formation and evolution of axion miniclusters and for the direct detection of axion dark matter on Earth. Finally we discuss the cosmology of axion models with N > 1 in which the domain wall problem is solved by introducing a small U P Q (1) breaking interaction. We find that in this case the walls decay into gravitational waves.PACS numbers:14.80. Mz,95.30.Cq

This paper revisits the problem of the string decay contribution to the axion cosmological energy density. We show that this contribution is proportional to the average relative increase when axion strings decay of a certain quantity N ax which we define. We carry out numerical simulations of the evolution and decay of circular and non-circular string loops, of bent strings with ends held fixed, and of vortex-antivortex pairs in two dimensions. In the case of string loops and of vortex-antivortex pairs, N ax decreases by approximately 20%. In the case of bent strings, N ax remains constant or increases slightly. Our results imply that the string decay contribution to the axion energy density is of the same order of magnitude as the well-understood contribution from vacuum realignment.

We investigate the cosmological constraints on exotic stable matter states which arise in realistic free fermionic superstring models. These states appear in the superstring models due to a "Wilson-line" breaking of the unifying nonAbelian gauge symmetry. In the models that we consider the unifying SO(10) gauge symmetry is broken at the string level to SO(6) × SO(4), SU (5) × U (1) or SU (3) × SU (2) × U (1) 2 . The exotic matter states are classified according to the patterns of the SO(10) symmetry breaking. In SO(6) × SO(4) and SU (5) × U (1) type models one obtains fractionally charged states with Q e.m. = ±1/2. In SU (3) × SU (2) × U (1) 2 type models one also obtains states with the regular charges under the Standard Model gauge group but with "fractional" charges under the U (1) Z ′ symmetry. These states include down-like color triplets and electroweak doublets, as well as states which are Standard Model singlets. By analyzing the renormalizable and nonrenormalizable terms of the superpotential in a specific superstring model, we show that these exotic states can be stable. We investigate the cosmological constraints on the masses and relic density of the exotic states. We propose that, while the abundance and the masses of the fractionally charged states are highly constrained, the Standard Model -like states, and in particular the Standard Model singlet, are good

Nuclear Physics, Section B 477 (1996) 65-104. doi:10.1016/0550-3213(96)00371-9Received by publisher: 1996-05-29Harvest Date: 2016-01-04 12:22:36DOI: 10.1016/0550-3213(96)00371-9Page Range: 65-10

We study the effect of the late decaying saxino (the scalar superpartner of the axion) and find out that there is a possible dark matter solution from a class of supersymmetric extensions of the invisible axion model. In this class of models, the saxino which decays into two axions acts as the late decaying . If we stick to the theoretical prejudice, Ω 0 = 1, we have at least two serious problems. First, we cannot fit the power spectrum curve even in the MDM model. Second, it would be inconsistent with the estimated lower bound on the age of the universe from the oldest globular clusters. An Ω 0 = 1 universe has an age of only t 0 = 6.5h −1 Gyr, giving t 0 < 9.3 Gyr for h > 0.7. On the other hand, the observed value of the age of the oldest globular cluster is around 15 Gyr [4]. Still there are a few possible way to relax this bound, but it seems not to be less than 11 Gyr [5]. Assuming that the data from the globular cluster can be relaxed by some mechanism and considering error bars in Hubble constant observations, h = 0.6 would be marginally allowed. However, the standard MDM model is inconsistent with large scale structure data even for h = 0.6.