Документ взят из кэша поисковой машины. Адрес оригинального документа : http://star.arm.ac.uk/~csj/papers/conference/IAUC187_BILyn.ps Дата изменения: Sat Jun 22 13:03:59 2002 Дата индексирования: Tue Oct 2 10:26:26 2012 Кодировка: Поисковые слова: http astrokuban.info astrokuban

**TITLE**
ASP Conference Series, Vol. **VOLUME**, **PUBLICATION YEAR**
**EDITORS**
The Strange Story of BI Lyncis: an Extremely
Helium-Rich \Subdwarf" with a Cool Companion
C. Simon Jeery, R. Aznar Cuadrado
Armagh Observatory, College Hill, Armagh BT61 9DG, Northern
Ireland
Abstract. BI Lyn = PG 0900+400 was previously classied as an evolved
binary system containing either a hot subdwarf or white dwarf and a thick
accretion disk. We have analyzed new intermediate dispersion spectra
and demonstrated that both views are wrong. BI Lyn is a highly-evolved
system containing a luminous low-mass helium star and a rapidly rotat-
ing G-type giant. It may be closely related to the evolved sdO+G binary
HD128220.
1. Introduction
The most unusual stars are frequently to be found in binary systems and to have
properties that are, supercially at least, unique. Sometimes this may be a result
of having too much data, sometimes too little. Another consequence of having
too little data is that an understandable misinterpretation of low-quality data
can grow into a full-blown myth in which a star takes on a persona quite divorced
from its true nature. In a recent paper (Jeery & Aznar Cuadrado 2001), one
such impostor, BI Lyncis { allegedly a \nova-like" cataclysmic variable, was
unmasked. This poster paper summarises the main points of its denouement as
a low-mass supergiant binary with an extremely evolved hot component.
2. The story so far
The star PG 0900+400 was identied in the Palomar-Green survey as a sub-
dwarf B (sdB) star with an infrared excess (Ferguson, Green & Liebert 1984),
although the overall ux distribution was notably redder than seen in other com-
posite systems. As such it was proposed to be a potential cataclysmic variable
(CV) progenitor. Photometric monitoring subsequently identied a 0.34d pe-
riod (Lipunova & Shugarov 1990, 1991, Kuczawska, Mikolajewski & Kirejczyk,
1993), from which the designation \nova-like" CV seems to have arisen (Downes,
Webbink & Shara 1997). This assignment may have been compounded by the
presence of a P Cygni type line prole at H, with a variable absorption com-
ponent (Wade & Potter 1995). Most recently, it was suspected to be a velocity
variable (Orosz, Wade & Harlow 1997). Successive spectral decompositions in-
dicated that the components were a subdwarf B star and a K3 dwarf (Ferguson
et al. 1984, Orosz et al. 1997, Liu & Hu 2000), perhaps because several similar
1

2 Jeery & Aznar Cuadrado
Figure 1. Ultraviolet and visual spectrophotometry of BI Lyn (his-
togram) together with the best tting theoretical ux distribution
(polyline and horizontal bars). The latter represents the sum of
two model atmospheres (dashed lines) with T e:1 = 28 600K,  1 =
0:55  10 11 rad, T e:2 = 5 840K and  2 = 4:09  10 11 rad. Interstellar
reddening is negligible.
such systems were well known. However reports of strong Hei lines (Orosz et al.
1997, Liu & Hu 2000) had not triggered any particular alarm bells.
Our own encounter with this star was almost accidental. We had undertaken
a spectroscopic survey of sdB stars with composite spectra (Aznar Cuadrado &
Jeery 2002). PG 0900+400 featured as a target but, in preparing the survey
sample, we had omitted a complete literature survey for all potential targets.
Had we encountered the SIMBAD designation \V* BI Lyn { Nova-like Star",
our target selection might have been dierent.
3. Observations
Moderate-resolution blue spectra of a large number of sdB stars were obtained
with the Isaac Newton Telescope of the La Palma Observatory. The spectrum
of PG 0900+400 was interesting because, instead of showing broad Balmer lines
and not much else as anticipated, it had a very line-rich spectrum. Initially it was
thought to be the spectrum of a G star and possibly a wrong target identication.
Closer inspection revealed that the strongest lines were due to Hei, with weaker
and quite narrow Balmer lines. It was more recognizable as an extreme helium
star than as a subdwarf B star. The spectrum also contained evidence of a

BI Lyn 3
G-band, calcium H+K lines and the near infrared calcium triplet { concrete
evidence for the presence of a late-type companion (Jeery & Pollacco 1988).
Subsequently, higher-resolution spectra were obtained with the William Herschel
Telescope in order to measure the atmospheric parameters more accurately. To
supplement the spectroscopic data, published photometry (Ferguson et al. 1984,
Lipinova & Shugarov 1991) and IUE images provided a description of the overall
ux distribution from 1100 { 9000  A.
Table 1. Atmospheric parameters for BI Lyn.
Star 1  2 
Spectrophotometry (ffit)
EB V 0.00 0.02
T e 28.6 1.0 5.84 0.96 kK
 0.55 0.01 4.09 0.10 10 11 rad
v t 5 a 2 a kms 1
R=R 1 1 7.44 0.03
Spectroscopy (sfit)
(T e 30.1 0.01 kK) b
T e 28.6 a 5.84 a kK
log g 3.6 0.1 3.2 0.3 (cgs)
n He 0.95 0.01 0.1 a
[ Fe ] 0.0 a 0.0 a
v sin i 0 a 120 20 kms 1
v t
5 a 2 a kms 1
R=R 1 1 4.9 0.5
a: assumed value
b: free solution not used
4. Atmospheric parameters
We rst carried out a reanalysis of the UV and optical ux distribution to
measure the eective temperatures (T e ) and angular diameters () of both
stars (Fig. 1). For this we used the automatic ux tting package ffit and a
combination of hydrogen-decient model atmospheres for hot stars and Kurucz
model atmospheres for cool stars (cf. Aznar Cuadrado & Jeery 2001).
The new INT and WHT intermediate resolution spectra of BI Lyn in the
blue (Fig. 2) and around H and the infrared calcium triplet were subsequently
used to carry out a simultaneous spectral analysis of both stars using the auto-
matic tting package sfit. Parameters to be solved for included T e (although
only that obtained from spectrophotometry was ultimately used), surface gravity
(log g), relative abundance of helium by number (n He ), logarithmic mass fraction
of metals relative to solar ([ Fe ]), rotational velocity, (v sin i) and relative radii
(R=R 1 ). The results of both analyses are given in Table 1.

4 Jeery & Aznar Cuadrado
Figure 2. Normalized blue spectrum of BI Lyn (bottom: d) together
with a best t composite model spectrum (c) formed by adding models
with a) T e:2 = 5 840K, log g 2 = 3:2, [Fe=H] 2 = 0:00 (top) and b)
T e:1 = 28 600K, log g 1 = 3:64, n He:1 = 0:95 assuming that the relative
radii R 2 =R 1 = 4:9. The model spectra have been velocity shifted and
degraded to match the observed spectral resolution (1  A).

BI Lyn 5
The analysis of the hot star based on the WHT blue spectrum yields a
hydrogen abundance  1% by number, T e  28 600K and log g  3:6, making
the primary a helium giant. Assuming a solar composition for the cool star,
its surface gravity is estimated from the CaT lines to be log g  3:2 and hence
also a giant. The two methods (ffit and sfit) give a radius ratio for the two
stars R 2 =R 1 between 7.4 and 4.9. Simply estimating M 1 = 0:5M enables all
remaining system dimensions to be evaluated, including estimates for distance,
luminosities and masses.
If it is assumed that the helium star mass is  0:5 M , and we adopt the
radius ratio R 2 =R 1  4:9  0:5 from spectrophotometry, then we nd radii of
1:9  0:3 and 9:1  1:9 R and luminosities (log L=L ) 3:3  0:1 and 1:9  0:1
for the hot and cool stars, respectively. The cool star mass is more problematic;
from the surface gravity we obtain M 2 = 4:8  1:8 M , a lower limit would be
 1:0  0:4 M .
5. Conclusions
Contrary to previous assumptions, BI Lyn does not contain a hot subdwarf
or white dwarf, nor is it a nova-like variable. The hot star is a luminous hot
hydrogen-decient star with a probable mass around 0:5M . Its luminosity
suggests that it lies on a post-AGB evolution track. The cool star is a giant
with a mass > 1M . Of particular interest is the fact that it is rotating rapidly
(v sin i = 120  20 kms 1 ).
The assumed mass is partially a result of the implied distance (> 5 kpc)
and consequent position in the galaxy. At a relatively high galactic latitude, a
signicantly larger total system mass is improbable from considerations of both
the secondary spectral type and the unlikely location for a system with such a
high mass.
It has been pointed out that the highly evolved star HD128220 contains a O-
type subdwarf and a rapidly-rotating G-type primary, probably spun up by mass
exchange (Howarth & Heber 1990). Both BI Lyn and HD128220 thus consist
of a hot post-AGB star and a more massive rapidly-rotating cool companion.
Signicantly, our estimates for the masss and luminosities of both components
in BI Lyn agree well with the estimates given by Howarth & Heber (1990) for
HD128220. The orbital period of HD128220 is 871 d. Similarly, we anticipate a
long period and relatively small velocity amplitudes (  < 30 km s 1 ) for BI Lyn.
The hydrogen deciency of the hot star is probably the result of at least one
common-envelope phase during which the outer envelope was entirely removed
or partially transferred to the cool companion. The helium star must at one
time have been substantially more massive than it is at present. Howarth &
Heber (1990) argue for an initial mass of 2 M for the sdO star in HD128220.
After losing some 70% of the original primary mass and inverting the original
mass ratio, the orbital separation would eventually increase towards the end of
the mass-transfer phase until the stars were no longer in contact. Having lost
its outer envelope, the primary can no longer sustain shell hydrogen burning
and so enters the post-AGB phase. BI Lyn and HD128220 are thus both long-
period binaries which will ultimately comprise a white dwarf and a G star. The
latter is eectively a \blue straggler", having accreted a substantial fraction

6 Jeery & Aznar Cuadrado
of the original primary during the mass transfer or common-envelope phase of
evolution.
There remains one crucial dierence between BI Lyn and HD128220, namely
the surface hydrogen abundance. HD128220 has a helium to hydrogen ratio
of 0.3 (Rauch 1993), compared with 20 for BI Lyn. If the two systems are
genuinely comparable, then the mechanism that terminates mass transfer and
the AGB phase must be able to produce a range of surface helium abundances.
An important clue will be provided by measuring the orbital period and velocities
for BI Lyn.
The previously-established 0.34d light variations are likely to be due to
pulsations in the helium star. By analogy with the hydrogen-decient binary
 Sgr, previously established H P-Cygni prole variations may be due to the
orbital motion of a supersonic jet between the components.
Further observations are required to determine the orbital period and mass
ratio, to verify the pulsation hypothesis, to correlate the H behaviour with
orbital phase and to establish the surface composition of both stars with greater
precision.
Acknowledgments. This research is based on observations obtained with
the Isaac Newton and William Herschel Telescopes, and on INES data from the
IUE satellite. This research has made use of the SIMBAD database, operated
at CDS, Strasbourg, France. CSJ is indebted to the Royal Society, London, and
the International Astronomical Union for awarding travel grants to attend IAU
Colloquium 187.
References
Aznar Cuadrado, R., & Jeery, C.S. 2001, A&A, 368, 994
Aznar Cuadrado, R., & Jeery, C.S. 2002, A&A, 385, 131
Downes, R., Webbink, R.F., & Shara, M.M. 1997, PASP, 109 345
Ferguson, D.H., Green, R.F., & Liebert, J. 1984, ApJ, 287, 320
Howarth, I.D., & Heber, U. 1990, PASP, 102, 912
Jeery, C.S., & Pollacco, D. 1998, MNRAS, 298, 179
Jeery, C.S., & Aznar Cuadrado, R. 2001, A&A, 378, 936
Kuczawska, E., Mikolajewski, M., & Kirejczyk K. 1993, IBVS, 3845
Lipunova, N.A., & Shugarov, S.Yu. 1990, JAVSO, 19, 40
Lipunova, N.A., & Shugarov, S.Yu. 1991, IBVS, 3580
Liu, W., & Hu, J.Y. 2000, ApJS, 128, 387
Orosz, J.A., Wade, R.A., & Harlow J.J.B. 1997, AJ, 114, 317
Rauch, T. 1993, A&A, 276, 171
Wade, R.A., & Potter, D. 1995, BAAS, 27Q.877