Multizone models of Type I X-ray bursts are presented that use an adaptive nuclear reaction network of unprecedented size, up to 1300 isotopes, for energy generation and include the most recent measurements and estimates of critical nuclear physics. Convection and radiation transport are included in calculations that carefully follow the changing composition in the accreted layer, both during the bursts themselves and in their ashes. Sequences of bursts, up to 15 in one case, are followed for two choices of accretion rate and metallicity, up to the point at which a limit cycle equilibrium is established. ForṀ ¼ 1:75 Â 10 À9 M yr À1 (andṀ ¼ 3:5 Â 10 À10 M yr À1 , for low metallicity), combined hydrogen-helium flashes occur. These bursts have light curves with slow rise times (seconds) and long tails. The rise times, shapes, and tails of these light curves are sensitive to the efficiency of nuclear burning at various waiting points along the rp-process path, and these sensitivities are explored. Each displays ''compositional inertia'' in that its properties are sensitive to the fact that accretion occurs onto the ashes of previous bursts that contain leftover hydrogen, helium, and CNO nuclei. This acts to reduce the sensitivity of burst properties to metallicity. Only the first anomalous burst in one model produces nuclei as heavy as A ¼ 100. For the present choice of nuclear physics and accretion rates, other bursts and models make chiefly nuclei with A % 64. The amount of carbon remaining after hydrogen-helium bursts is typically P1% by mass and decreases further as the ashes are periodically heated by subsequent bursts. Foṙ M ¼ 3:5 Â 10 À10 M yr À1 and solar metallicity, bursts are ignited in a hydrogen-free helium layer. At the base of this layer, up to 90% of the helium has already burned to carbon prior to the unstable ignition of the helium shell. These helium-ignited bursts have (1) briefer, brighter light curves with shorter tails, (2) very rapid rise times (<0.1 s), and (3) ashes lighter than the iron group.

Explosive hydrogen burning in type I X-ray bursts (XRBs) comprise charged particle reactions creating isotopes with masses up to A ∼ 100. Since charged particle reactions in a stellar environment are very temperature sensitive, we use a realistic time-dependent general relativistic and self-consistent model of type I x-ray bursts to provide accurate values of the burst temperatures and densities. This allows a detailed and accurate time-dependent identification of the reaction flow from the surface layers through the convective region and the ignition region to the neutron star ocean. Using this, we determine the relative importance of specific nuclear reactions in the X-ray burst.

One of the two breakout reactions from the hot CNO (HCNO) cycle is 15 O(,) 19 Ne, which at low temperatures depends strongly on the resonance strength of the 4.033 MeV state in 19 Ne. An experimental upper limit has been placed on its strength, but the lower limit on the resonance strength and therefore the astrophysical reaction rate is unconstrained experimentally. However, this breakout reaction is crucial to the thermonuclear runaway that causes type I X-ray bursts on accreting neutron stars. In this paper we exploit astronomical observations in an attempt to constrain the relevant nuclear physics and deduce a lower limit on the reaction rate. Our sensitivity study implies that if the rate were sufficiently small, accreting material would burn stably without bursts. The existence of type I X-ray bursts and superbursts consequently suggests a lower limit on the 15 O(,) 19 Ne reaction rate at low temperatures.

Type I X-ray bursts with a double peak in the bolometric luminosity have been observed from several sources. The separation between the two peaks are on the order of a few seconds. We propose a nuclear waiting point impedance in the thermonuclear reaction flow to explain these observations. Nuclear structure information suggests the potential waiting points: 26 Mg, 26 Si, 30 S, and 34 Ar, which arise in conditions, where a further reaction flow has to await a β + -decay, because the (α, p)-reaction is too weak to overcome the target Coulomb-barrier and the (p, γ)reaction is quenched by photo-disintegration at the burst temperature. The conclusion is that the effects of the experimentally unknown 30 S(α, p) 33 Cl and 34 Ar(α, p) 37 K might be directly visible in the observation of X-ray burst light curves.