Issue 
A&A
Volume 576, April 2015



Article Number  A118  
Number of page(s)  18  
Section  Stellar structure and evolution  
DOI  https://doi.org/10.1051/00046361/201424739  
Published online  16 April 2015 
Online material
Appendix A: Bestfitting models to the observed CEMPs stars in our sample
In this section we show the bestfitting models to the observed abundances of all CEMPs stars in our sample.
Fig. A.1
Points with error bars: as Fig. 4 for BD+ 04°2466. Solid line: bestfitting model computed with model set A (Table 2). The combined abundance of carbon and nitrogen is shown in the right panel. 

Open with DEXTER 
Fig. A.2
Star CS22948–027. Solid line and points: as Fig. A.1. In the best model (with model set B) the secondary star accretes ΔM_{acc} = 0.27 M_{⊙}. With model set A (dashed line) the secondary star accretes ΔM_{acc} = 0.12 M_{⊙} and therefore the material is more strongly diluted. 

Open with DEXTER 
Fig. A.3
Solid line and points: as Fig. A.1 for CS22956–028. The bestfitting model assumes that nonconvective mixing processes, such as thermohaline mixing, are inefficient and therefore the accreted material remains on the surface of the star. Dashed line: alternative model with efficient thermohaline mixing (χ^{2} = 51.8). 

Open with DEXTER 
Fig. A.4
Points with error bars: as in Fig. A.1 for CS29497–034. Solid line: bestfitting model computed with set B adopting the default metallicity Z = 10^{4}. is dominated by the discrepancy in the abundance of europium (about 1 dex). If we exclude europium from the fit we find and . Dashed line: alternative model computed with set B and reduced metallicity, Z = 2 × 10^{5}, that corresponds to the observed iron abundance [Fe/H] = −2.96. 

Open with DEXTER 
Fig. A.5
As Fig. A.1 for CS29509–027. The abundance of other elemens, such as N, Na, Mg and Pb, is necessary to better constrain the initial primary mass. 

Open with DEXTER 
Fig. A.6
As Fig. A.1 for HD 198269. The bestfitting model (solid line) is computed adopting model set B. The abundances determined by Van Eck et al. (2003) and Vanture (1992b) are represented as plus signs and filled circles, respectively. 

Open with DEXTER 
Fig. A.7
As Fig. A.1 for HD 201626. The bestfitting model to the observed abundances is found with model set B. 

Open with DEXTER 
Fig. A.8
As Fig. A.1 for CEMPs/r star HD 224959. The abundances determined by Van Eck et al. (2003) and Vanture (1992b) are represented as plus signs and filled circles, respectively. 

Open with DEXTER 
Fig. A.9
Points with error bars: as Fig. A.1 for HE0024–2523. Solid line: bestfitting model found with model set B and assuming a low efficiency for commonenvelope ejection, α_{CE} = 0.03, which is required to reproduce the observed period of P_{orb} = 3.14 days in a binary scenario. Dashed line: alternative scenario in which the HE0024–2523 was initially part of a hierarchical triple system. An intermediatemass primary star was in a wide orbit around a close binary. During its AGB phase the primary star pollutes the inner binary. Subsequently, the secondary star overfills its Roche lobe, the system enters a common envelope the ejection of which (modelled with α_{CE} = 1.0) requires the orbit to shrink to the observed period. 

Open with DEXTER 
Fig. A.10
As Fig. A.1 for star HE0507–1430. The bestfit model is found with model set B. 

Open with DEXTER 
Fig. A.11
As in Fig. A.1 for star LP625–44. The bestfit model is found with model set B. 

Open with DEXTER 
Appendix B: Twodimensional confidence intervals
Figure B.1a shows the twodimensional confidence regions determined for the initial orbital period, P_{i}, and primary mass, M_{1,i}, of model star CS29497–034. The onedimensional confidence intervals of these two parameters are shown in Figs. B.1b and c, respectively, in which the same symbols as in Fig. 5 are used. Figure B.2 is the same as Fig. B.1 for star HD 201626. To compute the twodimensional confidence intervals of two input parameters p and p′ we follow the same procedure as described in Sect. 4.3, but we fix a pair of values (p,p′) and we let the other two parameters vary. In two dimensions the confidence levels of 68.3%,95.4% and 99.7% correspond respectively to the thresholds Δχ^{2} = 2.30, 6.17, 11.8 (Press et al. 1989). These thresholds are shown in Figs. B.1a and B.2a as solid, dashed and dotted lines, respectively.
The twodimensional confidence intervals in Fig. B.1a indicate that there is correlation between the initial orbital period and primary mass of the model systems. The more massive the primary star is, the longer needs to be the initial period to minimise χ^{2}. In Fig. B.2a the correlation between the two parameters in the models of star HD 201626 is even more clear. This correlation indicates that although the confidence range of M_{1,i} is rather large, as shown in Table 4, for each primary mass in our grid there is essentially only one orbital period at which the model reproduces the observations.
Fig. B.1
Confidence intervals of the input parameters for model star CS29497–034. Panels b) and c): onedimensional confidence intervals of P_{i} and M_{1,i}, respectively. Longdashed lines indicate the thresholds Δχ^{2} = 1, 4, 9. Panel a): twodimensional confidence intervals. The thresholds Δχ^{2} = 2.30, 6.17, 11.8 are represented as solid, dashed and dotted lines, respectively. The blue plus sign indicates the best model. 

Open with DEXTER 
Fig. B.2
Confidence intervals of the input parameters for model star HD 201626. The symbols are as in Fig. B.1. 

Open with DEXTER 
© ESO, 2015
Current usage metrics show cumulative count of Article Views (fulltext article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 4896 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.