4 Accreting Millisecond X-ray Pulsars
183
fractional amplitude of the first overtone and the X-ray flux (something very similar
was also observed in XTE J1807-294; [259]), suggesting that the secondary hot spot
becomes visible or more prominent as the accretion process becomes less intense.
This may be explained if the inner disk radius moves in or out with respect to the
accretion rate, rendering the secondary hot-spot visible [123, 142]. If the inner
disk radius r in is assumed to vary at different outburst stages, then the pulse profile
changes of SAX J1808.4-3658 can be explained in this way [279]. Variations in the
disk-magnetosphere coupling may also explain explain some of the pulse shape and
phase variations [159]. Calculation of the light-curve of accreting NSs in global
3D-MHD simulations showed that if an octupolar magnetic field dominates the
field configuration, double peaked pulse profiles can be reproduced [199]. Similar
pulse shape variations were seen to correlate with the outburst stage of Swift
J1756.9-2508 [259] IGR J17511-3057 [142] and to some extent XTE J1807-294
[249]. However, to date it is still unclear why pulse profiles change abruptly and
unpredictably in some sources during certain outbursts but at other times have
limited variability.
4.6 Long Term Evolution and Pulse Formation Process
If AMXP pulsations are observed in different outbursts one can follow the longterm spin and orbital evolution. To date only five AMXPs have been monitored with
high time resolution instruments in different outbursts: SAX J1808.4-3658, IGR
J00291+5934, XTE J1751-305, Swift J1756.9-2508 and NGC6440 X-2 (although
with relatively low S/N and short outburst duration, it is difficult to constrain the
long-term evolution of the latter two).
4.6.1 Specific Sources
SAX J1808.4-3658 The best constrained AMXP is SAX J1808.4-3658, for which
secular spin evolution and orbital period variation have now been measured over a
13 year baseline [123, 125, 262]. As discussed previously, accretion torques have a
small effect on the spin of SAX J1808.4-3658 and have never been conclusively
detected during any of the six outbursts monitored by RXTE. However, if one
compares the constant spin frequency measured in each outburst, SAX J1808.43658 is clearly spinning down at a constant rate (bottom panel in Fig. 4.8; [262]).
The stability of the spin-down rate suggests that its origin is unrelated to propeller
onset, since spin-down would then depend on the amount of mass propelled and on
the duration of this phase, which can vary from outburst to outburst. Loss of angular
momentum via emission of gravitational waves was also considered [123], and is
interesting since this might also balance accretion torques, explaining the lack of
measurable spin-up during the outbursts. However, fine tuning is also required in
183
fractional amplitude of the first overtone and the X-ray flux (something very similar
was also observed in XTE J1807-294; [259]), suggesting that the secondary hot spot
becomes visible or more prominent as the accretion process becomes less intense.
This may be explained if the inner disk radius moves in or out with respect to the
accretion rate, rendering the secondary hot-spot visible [123, 142]. If the inner
disk radius r in is assumed to vary at different outburst stages, then the pulse profile
changes of SAX J1808.4-3658 can be explained in this way [279]. Variations in the
disk-magnetosphere coupling may also explain explain some of the pulse shape and
phase variations [159]. Calculation of the light-curve of accreting NSs in global
3D-MHD simulations showed that if an octupolar magnetic field dominates the
field configuration, double peaked pulse profiles can be reproduced [199]. Similar
pulse shape variations were seen to correlate with the outburst stage of Swift
J1756.9-2508 [259] IGR J17511-3057 [142] and to some extent XTE J1807-294
[249]. However, to date it is still unclear why pulse profiles change abruptly and
unpredictably in some sources during certain outbursts but at other times have
limited variability.
4.6 Long Term Evolution and Pulse Formation Process
If AMXP pulsations are observed in different outbursts one can follow the longterm spin and orbital evolution. To date only five AMXPs have been monitored with
high time resolution instruments in different outbursts: SAX J1808.4-3658, IGR
J00291+5934, XTE J1751-305, Swift J1756.9-2508 and NGC6440 X-2 (although
with relatively low S/N and short outburst duration, it is difficult to constrain the
long-term evolution of the latter two).
4.6.1 Specific Sources
SAX J1808.4-3658 The best constrained AMXP is SAX J1808.4-3658, for which
secular spin evolution and orbital period variation have now been measured over a
13 year baseline [123, 125, 262]. As discussed previously, accretion torques have a
small effect on the spin of SAX J1808.4-3658 and have never been conclusively
detected during any of the six outbursts monitored by RXTE. However, if one
compares the constant spin frequency measured in each outburst, SAX J1808.43658 is clearly spinning down at a constant rate (bottom panel in Fig. 4.8; [262]).
The stability of the spin-down rate suggests that its origin is unrelated to propeller
onset, since spin-down would then depend on the amount of mass propelled and on
the duration of this phase, which can vary from outburst to outburst. Loss of angular
momentum via emission of gravitational waves was also considered [123], and is
interesting since this might also balance accretion torques, explaining the lack of
measurable spin-up during the outbursts. However, fine tuning is also required in