5 Thermonuclear X-ray Bursts
249
At the onset of a burst, the flame travels a longer path than what can be fully
resolved by current simulations. Two-dimensional simulations that include the
radial and latitudinal (or longitudinal) directions, adopting zones that are elongated
in the latter dimension. The zones are not fine enough to resolve the small-scale
turbulence during the convective stage at the burst onset, but they are sufficient
for modeling the larger scale mechanisms that drive the flame propagation. These
simulations find that initially, the flame must be confined by the Coriolis force and/or
magnetic tension in order to prevent it from dying out [21–23, 166]. Subsequently, a
combination of the Coriolis force, magnetic tension-induced friction in the sheared
burning front, and conduction causes the flame to spread across the surface.
5.9 Nuclear Experimental Physics
A significant fraction of the research effort directed at thermonuclear bursts over the
last few decades has been in the area of nuclear experimental physics. The nuclear
reactions that may take place in bursts, where the temperature can reach 10 8 –10 9 K,
were first explored by Wallace and Woosley [192]. The most important reactions
are the β-limited CNO cycle, in systems that accrete hydrogen, and the triple-α
reaction, which fuses helium to carbon.
The triple-α reaction first produces carbon, which (via the CNO cycle) catalyses
the burning of hydrogen. CNO breakout reactions including 15 O(α, γ ) 19 Ne then
lead to a series of (α, p) and (p, γ ) reactions (the αp-process), feeding into
the rapid-proton, or rp-process. The rp-process was identified as a key reaction
chain that could produce elements far heavier than any of the precursor reactions.
However, experimental data for the neutron-deficient nuclei involved was lacking
for many years, forcing modelers to rely on theoretical calculations for masses
and rates [158]. At the same time, it was initially unclear how far the rp-process
reactions extended. Early simulations were based on limited reaction networks,
under the assumption that further proton captures beyond 56 Ni could be neglected.
Such simulations also tended to find non-negligible reaction products accumulating
at the limits of the reaction rate networks. A simulation study using a one-zone
model determined the ultimate limit of rp-process burning in the Sn-Sb-Te cycle
[159] (see also [106]). This result implies that rp-process reaction products are
limited to atomic number Z ≤ 54, and that the ashes forming the crust are made up
of nuclei lighter than mass number A ≈ 107.
With the maximum extent of the rp-process burning now constrained, the
goals of experimenters became to improve the mass measurements for protonrich nuclei (e.g. [157]), as well as determining the key reactions that might
influence the properties of the bursts. The experimental studies informed, and were
further motivated by, development of more accurate numerical models of bursts, as
described in Sect. 5.8. Simulations enabled a very detailed picture of the nuclear
burning processes during a burst to emerge (e.g. [46]), although the fidelity of this
picture remains uncertain.
249
At the onset of a burst, the flame travels a longer path than what can be fully
resolved by current simulations. Two-dimensional simulations that include the
radial and latitudinal (or longitudinal) directions, adopting zones that are elongated
in the latter dimension. The zones are not fine enough to resolve the small-scale
turbulence during the convective stage at the burst onset, but they are sufficient
for modeling the larger scale mechanisms that drive the flame propagation. These
simulations find that initially, the flame must be confined by the Coriolis force and/or
magnetic tension in order to prevent it from dying out [21–23, 166]. Subsequently, a
combination of the Coriolis force, magnetic tension-induced friction in the sheared
burning front, and conduction causes the flame to spread across the surface.
5.9 Nuclear Experimental Physics
A significant fraction of the research effort directed at thermonuclear bursts over the
last few decades has been in the area of nuclear experimental physics. The nuclear
reactions that may take place in bursts, where the temperature can reach 10 8 –10 9 K,
were first explored by Wallace and Woosley [192]. The most important reactions
are the β-limited CNO cycle, in systems that accrete hydrogen, and the triple-α
reaction, which fuses helium to carbon.
The triple-α reaction first produces carbon, which (via the CNO cycle) catalyses
the burning of hydrogen. CNO breakout reactions including 15 O(α, γ ) 19 Ne then
lead to a series of (α, p) and (p, γ ) reactions (the αp-process), feeding into
the rapid-proton, or rp-process. The rp-process was identified as a key reaction
chain that could produce elements far heavier than any of the precursor reactions.
However, experimental data for the neutron-deficient nuclei involved was lacking
for many years, forcing modelers to rely on theoretical calculations for masses
and rates [158]. At the same time, it was initially unclear how far the rp-process
reactions extended. Early simulations were based on limited reaction networks,
under the assumption that further proton captures beyond 56 Ni could be neglected.
Such simulations also tended to find non-negligible reaction products accumulating
at the limits of the reaction rate networks. A simulation study using a one-zone
model determined the ultimate limit of rp-process burning in the Sn-Sb-Te cycle
[159] (see also [106]). This result implies that rp-process reaction products are
limited to atomic number Z ≤ 54, and that the ashes forming the crust are made up
of nuclei lighter than mass number A ≈ 107.
With the maximum extent of the rp-process burning now constrained, the
goals of experimenters became to improve the mass measurements for protonrich nuclei (e.g. [157]), as well as determining the key reactions that might
influence the properties of the bursts. The experimental studies informed, and were
further motivated by, development of more accurate numerical models of bursts, as
described in Sect. 5.8. Simulations enabled a very detailed picture of the nuclear
burning processes during a burst to emerge (e.g. [46]), although the fidelity of this
picture remains uncertain.