Non-linear oscillator models for the X-ray bursting of the microquasar GRS 1915+105

The microquasar GRS 1915+105, exhibits a large variety of characteristic states, according to its luminosity, spectral state, and variability. The most interesting one is the so-called rho-state, whose light curve shows recurrent bursts. This paper presents a model based on Fitzhugh-Nagumo equations containing two variables: x, linked to the source photon luminosity L detected by the MECS, and y related to the mean photon energy. We aim at providing a simple mathematical framework composed by non-linear differential equations useful to predict the observed light curve and the energy lags for the rho-state and possibly other classes of the source. We studied the equilibrium state and the stability conditions of this system that includes one external parameter, J, that can be considered a function of the disk accretion rate. Our work is based on observations performed with the MECS on board BeppoSAX when the source was in rho and nu mode, respectively. The evolution of the mean count rate and photon energy were derived from a study of the trajectories in the count rate - photon energy plane. Assuming J constant, we found a solution that reproduces the x profile of the rho class bursts and, unexpectedly, we found that y exhibited a time modulation similar to that of the mean energy. Moreover, assuming a slowly modulated J the solutions for x quite similar to those observed in the nu class light curves is reproduced. According these results, the outer mass accretion rate is probably responsible for the state transitions, but within the rho-class it is constant. This finding makes stronger the heuristic meaning of the non-linear model and suggests a simple relation between the variable x and y. However, how a system of dynamical equations can be derived from the complex mathematical apparatus of accretion disks remains to be furtherly explored.

Effects of capillary reflection in the performance of the collimator of the Large Area Detector on board LOFT

The Large Observatory For X-ray Timing (LOFT) is one of the candidate missions selected by the European Space Agency for an initial assessment phase in the Cosmic Vision programme. It is proposed for the M3 launch slot and has broad scientific goals related to fast timing of astrophysical X-ray sources. LOFT will carry the Large Area Detector (LAD), as one of the two core science instruments, necessary to achieve the challenging objectives of the project. LAD is a collimated detector working in the energy range 2-50 keV with an effective area of approximately 10 m^2 at 8 keV. The instrument comprises an array of modules located on deployable panels. Lead-glass microchannel plate (MCP) collimators are located in front of the large-area Silicon Drift Detectors (SDD) to reduce the background contamination from off-axis resolved point sources and from the diffuse X-ray background. The inner walls of the microchannel plate pores reflect grazing incidence X-ray photons with a probability that depends on energy. In this paper, we present a study performed with an ad-hoc simulator of the effects of this capillary reflectivity on the overall instrument performance. The reflectivity is derived from a limited set of laboratory measurements, used to constrain the model. The measurements were taken using a prototype collimator whose thickness is similar to that adopted in the current baseline design proposed for the LAD. We find that the experimentally measured level of reflectivity of the pore inner walls enhances the off-axis transmission at low energies, producing an almost flat-top response. The resulting background increase due to the diffuse cosmic X-ray emission and sources within the field of view does not degrade the instrument sensitivity.

The complex time behaviour of the microquasar GRS 1915+105 in the rho-class observed with BeppoSAX. III: The hard X-ray delay and limit cycle mapping

The microquasar GRS1915+105 was observed by BeppoSAX in October 2000 for about ten days while the source was in rho-mode, which is characterized by a quasi-regular type I bursting activity. This paper presents a systematic analysis of the delay of the hard and soft X-ray emission at the burst peaks. The lag, also apparent from the comparison of the [1.7-3.4] keV light curves with those in the [6.8-10.2] keV range, is evaluated and studied as a function of time, spectral parameters, and flux. We apply the limit cycle mapping technique, using as independent variables the count rate and the mean photon rate. The results using this technique were also cross-checked using a more standard approach with the cross-correlation methods. Data are organized in runs, each relative to a continuous observation interval. The detected hard-soft delay changes in the course of the pointing from about 3 s to about 10 s and presents a clear correlation with the baseline count rate.