216
D. K. Galloway and L. Keek
at ∼30% ˙
M Edd [28, 53, 187]. Furthermore, close to this transition the observed
bursts have relatively long or irregular recurrence times with indications of stable
burning in-between bursts (high α-values, indicative of low Q nuc ). This bursting
regime does not match any of those in Table 5.1 (see also Sect. 5.1.2). To improve
the theoretical predictions, several processes have been investigated in recent years.
First, a “base heating” term is included for most models, which only describe
the outer layers of the neutron star. This additional flux, usually specified by the
parameter Q b (in units of MeV per accreted nucleon), models the heat flowing
from the underlying neutron star crust into the envelope. The corresponding base
luminosity, L b = ˙
MQ b , is thought to include contributions from pycnonuclear
and electron-capture reactions in the crust. Most simulations use a low value of
Q b = 0.1–0.15 MeV u −1 , stemming from an early prediction [66]. However, more
recent work indicates that the generated heat may be larger, up to Q b = 2 MeV u −1
[60, 67]. Additionally, a yet-unknown shallow heat source may increase Q b [18, 39].
The degree of heating from below the fuel layer Q b may be reduced by the
competing effect of neutrino cooling. Neutrino emissions from the core and crust
strongly depend on temperature (e.g., [32]), and are therefore higher at larger ˙
M:
Q b 1.0 MeV u −1 at 0.01 ˙
M Edd and Q b 0.1 MeV u −1 at ≥ 0.1 ˙
M Edd [32].
Recently, a new Urca neutrino cooling process in the outer crust has been predicted,
which further complicates estimates of Q b [161]. This parameter can affect the
accretion rate range in which the burst regimes of Sect. 5.1.1.4 occur; for example,
larger value of Q b places the transition between bursts and stable burning at a lower
˙
M [98, 206].
Second, rotational mixing may be induced through the accretion of matter onto
the neutron star, which includes the transfer of angular momentum. This mixing is
turbulent, and may involve a rotationally-induced magnetic field [98, 151]. Mixing
transports the accreted fuel more quickly down to the ignition depth, where it can
burn in a stable manner. Therefore, rotational mixing also reduces the ˙
M where the
stability of nuclear burning is predicted to change.
Third, under the influence of the strong surface gravity, heavier isotopes sink
faster than lighter ones, leading to gravitational separation. At low mass accretion
rates ( ˙
M 0.01 ˙
M Edd ) the burst recurrence times are sufficiently long for
sedimentation of the fuel layer to occur. When He and CNO separate from H, this
influences the ignition conditions of H-ignited bursts (regimes I and II in Table 5.1)
[149]. Deeper in the envelope, chemical separation may occur of carbon and heavier
isotopes, which influences the ignition conditions of superbursts [129, 130].
Additional factors that may be important include the ignition latitude
(Sect. 5.2.3), multi-zone and multi-dimensional effects (Sect. 5.8), and the accuracy
of nuclear physics data (Sect. 5.9). These additional effects lead to much larger
parameter-space for different ignition regimes than the ˙
M dependence alone. Only
a small part of this parameter space has been explored with numerical models at
this time.
D. K. Galloway and L. Keek
at ∼30% ˙
M Edd [28, 53, 187]. Furthermore, close to this transition the observed
bursts have relatively long or irregular recurrence times with indications of stable
burning in-between bursts (high α-values, indicative of low Q nuc ). This bursting
regime does not match any of those in Table 5.1 (see also Sect. 5.1.2). To improve
the theoretical predictions, several processes have been investigated in recent years.
First, a “base heating” term is included for most models, which only describe
the outer layers of the neutron star. This additional flux, usually specified by the
parameter Q b (in units of MeV per accreted nucleon), models the heat flowing
from the underlying neutron star crust into the envelope. The corresponding base
luminosity, L b = ˙
MQ b , is thought to include contributions from pycnonuclear
and electron-capture reactions in the crust. Most simulations use a low value of
Q b = 0.1–0.15 MeV u −1 , stemming from an early prediction [66]. However, more
recent work indicates that the generated heat may be larger, up to Q b = 2 MeV u −1
[60, 67]. Additionally, a yet-unknown shallow heat source may increase Q b [18, 39].
The degree of heating from below the fuel layer Q b may be reduced by the
competing effect of neutrino cooling. Neutrino emissions from the core and crust
strongly depend on temperature (e.g., [32]), and are therefore higher at larger ˙
M:
Q b 1.0 MeV u −1 at 0.01 ˙
M Edd and Q b 0.1 MeV u −1 at ≥ 0.1 ˙
M Edd [32].
Recently, a new Urca neutrino cooling process in the outer crust has been predicted,
which further complicates estimates of Q b [161]. This parameter can affect the
accretion rate range in which the burst regimes of Sect. 5.1.1.4 occur; for example,
larger value of Q b places the transition between bursts and stable burning at a lower
˙
M [98, 206].
Second, rotational mixing may be induced through the accretion of matter onto
the neutron star, which includes the transfer of angular momentum. This mixing is
turbulent, and may involve a rotationally-induced magnetic field [98, 151]. Mixing
transports the accreted fuel more quickly down to the ignition depth, where it can
burn in a stable manner. Therefore, rotational mixing also reduces the ˙
M where the
stability of nuclear burning is predicted to change.
Third, under the influence of the strong surface gravity, heavier isotopes sink
faster than lighter ones, leading to gravitational separation. At low mass accretion
rates ( ˙
M 0.01 ˙
M Edd ) the burst recurrence times are sufficiently long for
sedimentation of the fuel layer to occur. When He and CNO separate from H, this
influences the ignition conditions of H-ignited bursts (regimes I and II in Table 5.1)
[149]. Deeper in the envelope, chemical separation may occur of carbon and heavier
isotopes, which influences the ignition conditions of superbursts [129, 130].
Additional factors that may be important include the ignition latitude
(Sect. 5.2.3), multi-zone and multi-dimensional effects (Sect. 5.8), and the accuracy
of nuclear physics data (Sect. 5.9). These additional effects lead to much larger
parameter-space for different ignition regimes than the ˙
M dependence alone. Only
a small part of this parameter space has been explored with numerical models at
this time.