Most of fast radio bursts (FRB) do not show evidence for repetition, and suchnon-repeating FRBs may be produced at the time of a merger of binary neutronstars (BNS), provided that the BNS merger rate is close to the high end of thecurrently possible range. However, the merger environment is polluted bydynamical ejecta, which may prohibit the radio signal to propagate. We examinethis by using a general-relativistic simulation of a BNS merger, and show thatthe ejecta appears about 1 ms after the rotation speed of the merged starbecomes the maximum. Therefore there is a time window in which an FRB signalcan reach outside, and the short duration of non-repeating FRBs can beexplained by screening after ejecta formation. A fraction of BNS mergers mayleave a rapidly rotating and stable neutron star, and such objects may be theorigin of repeating FRBs like FRB 121102. We show that a merger remnant wouldappear as a repeating FRB in a time scale of about 1-10 yrs, and expectedproperties are consistent with the observations of FRB 121102. We construct anFRB rate evolution model including these two populations of repeating andnon-repeating FRBs from BNS mergers, and show that the detection rate ofrepeating FRBs relative to non-repeating ones rapidly increases with improvingsearch sensitivity. This may explain that the only repeating FRB 121102 wasdiscovered by the most sensitive FRB search with Arecibo. Several predictionsare made, including appearance of a repeating FRB 1-10 years after a BNS mergerthat is localized by gravitational wave and subsequent electromagneticradiation.
展开▼
