No Arabic abstract
We have monitored the pulse frequencies of the two soft gamma repeaters SGR 1806-20 and SGR 1900+14 through the beginning of year 2001 using primarily Rossi X-ray Timing Explorer Proportional Counter Array observations. In both sources, we observe large changes in the spin-down torque up to a factor of ~4, which persist for several months. Using long baseline phase-connected timing solutions as well as the overall frequency histories, we construct torque noise power spectra for each SGR. The power spectrum of each source is very red (power-law slope ~-3.5). The torque noise power levels are consistent with some accreting systems on time scales of ~1 year, yet the full power spectrum is much steeper in frequency than any known accreting source. To the best of our knowledge, torque noise power spectra with a comparably steep frequency dependence have only been seen in young, glitching radio pulsars (e.g. Vela). The observed changes in spin-down rate do not correlate with burst activity, therefore, the physical mechanisms behind each phenomenon are also likely unrelated. Within the context of the magnetar model, seismic activity cannot account for both the bursts and the long-term torque changes unless the seismically active regions are decoupled from one another.
In this short note I discuss the hypothesis that bursting activity of magnetars evolves in time analogously to the glitching activity of normal radio pulsars (i.e. sources are more active at smaller ages), and that the increase of the burst rate follows one of the laws established for glitching radio pulsars. If the activity of soft gamma repeaters decreases in time in the way similar to the evolution of core-quake glitches ($propto t^{5/2}$), then it is more probable to find the youngest soft gamma repeaters, but the energy of giant flares from these sources should be smaller than observed $10^{44}$ --$10^{46}$ ergs as the total energy stored in a magnetars magnetic field is not enough to support thousands of bursts similar to the prototype 5 March 1979 flare.
In this paper I will briefly review what are, in my view, the main contributions of BeppoSAX to the understanding of the class of sources known as Soft Gamma Repeaters. These enigmatic sources were firmly identified as steady pulsars just during the operating lifetime of BeppoSAX. All the instruments onboard BeppoSAX have at some level contributed in this field with specific observations, always allowing high quality - sometimes unprecedented - studies of the quiescent counterparts or the bursting behavior of these sources. I will try to stress the results that were uniquely achieved by BeppoSAX and identify their impact on the knowledge of the physics at work in these sources.
We argue that giant flares in SGRs can be associated to the core conversion of an isolated neutron star having a subcritical magnetic field $sim 10^{12}$ G and a fallback disk around it. We show that, in a timescale of $lesssim 10^5$ yrs, accretion from the fallback disk can increase the mass of the central object up to the critical mass for the conversion of the core of the star into quark matter. A small fraction of the neutrino-antineutrino emission from the just-converted quark-matter hot core annihilates into $e^+e^-$ pairs above the neutron star surface originating the gamma emission of the spike while the further cooling of the heated neutron star envelope originates the tail of the burst. We show that several characteristics of the giant flare of the SGR 1806-20 of 27 December 2004 (spike and tail energies, timescales, and spectra) can be explained by this mechanism.
Infrared observations of the environment of the two Soft Gamma-ray Repeaters (SGRs) with the best known locations on the sky show that they are associated to clusters of massive stars. Observations with ISO revealed that SGR 1806-20 is in a cluster of giant massive stars, still enshrouded in a dense cloud of gas and dust. SGR 1900+14 is at the edge of a similar cluster that was recently found hidden in the glare of a pair of M5 supergiant stars. Since none of the stars of these clusters has shown in the last years significant flux variations in the infrared, these two SGRs do not form bound binary systems with massive stars. SGR 1806-20 is at only ~ 0.4 pc, and SGR 1900+14 at ~ 0.8 pc from the centers of their parental star clusters. If these SGRs were born with typical neutron star runaway velocities of ~ 300 km/s, they are not older than a few 10$^{3}$ years. We propose that SGR 1806-20 and SGR 1900+14 are ideal laboratories to study the evolution of supernovae explosions inside interstellar bubbles produced by the strong winds that prevail in clusters of massive stars.
We present the results of a LIGO search for short-duration gravitational waves (GWs) associated with Soft Gamma Repeater (SGR) bursts. This is the first search sensitive to neutron star f-modes, usually considered the most efficient GW emitting modes. We find no evidence of GWs associated with any SGR burst in a sample consisting of the 27 Dec. 2004 giant flare from SGR 1806-20 and 190 lesser events from SGR 1806-20 and SGR 1900+14 which occurred during the first year of LIGOs fifth science run. GW strain upper limits and model-dependent GW emission energy upper limits are estimated for individual bursts using a variety of simulated waveforms. The unprecedented sensitivity of the detectors allows us to set the most stringent limits on transient GW amplitudes published to date. We find upper limit estimates on the model-dependent isotropic GW emission energies (at a nominal distance of 10 kpc) between 3x10^45 and 9x10^52 erg depending on waveform type, detector antenna factors and noise characteristics at the time of the burst. These upper limits are within the theoretically predicted range of some SGR models.