UZ Lib is a star whose true nature was misinterpreted for several decades.
Over 60 years ago Parenago (1931) found it to have light variations
and concluded that it was an RR Lyrae variable. From new photometry
Wiesniewski (1973) determined a period of 9.5 days and classified
it as a Lyrae-type system. Evans & Bopp (1974) proposed
that UZ Lib might be a spotted-dwarf flare star, a suggestion that
ultimately proved to be close to the mark. Hoffmann (1980) obtained
additional photometry and found it consistent with a period of 4.75 days.
Bopp et al. (1981) confirmed the likelyhood of the 4.75
day photometric period and determined a V-band amplitude of 0
35.
Recently, from V-band photometry spanning 25 years Strassmeier et al.
(1997) obtained a long-term average photometric period of
4.75226 days.
Bidelman (1954) reported the discovery of its CaII H and K emission, gave its spectral type as Kp, and listed it in a table of spectroscopic binaries, although no reference for its duplicity was given. Bopp & Stencel (1981) proposed that UZ Lib is an FK Comae-type star. This class of photometric variables consists of a very small group of chromospherically active single late-type giants that have extremely rapid rotation. The radial velocities of Woolley et al. (1981), however, indicated that the star is a binary. Bopp et al. (1984) obtained additional velocities and determined an orbital period of 4.7678 days.
Two new sets of velocities were obtained by the Vienna group. From March 1994 through April 1998, 49 observations were made at KPNO, while 9 additional velocities were obtained in May 1996 at ESO. Velocities of the March 1994 spectra have previously been published by Strassmeier (1996). Our velocities for those spectra are slightly different because of different assumed velocities for the standard stars.
For our KPNO data an initial set of orbital elements was determined with BISP and refined with SB1. Then the ESO velocities, given unit weight, were included in a new orbital solution. In a similar manner independent orbital solutions were determined for the Woolley et al. (1981) and the Bopp et al. (1984) velocities. From the resulting variance of the data set and the center-of-mass velocity, the Woolley et al. (1981) data were given weights of 0.2, and 10.0 kms-1 was added to each velocity. Results for the Bopp et al. (1984) velocities were 0.25 and -2.0 kms-1. Because of their small number and uncertain velocity offset, the velocities of Grewing et al. (1989) were not included in our analysis.
An SB1 solution with the four
velocity sets appropriately weighted, except for three velocities given
zero weight, was compared to a solution of our new velocities that
had the period fixed from the all-data solution. Since only the
error of the period was improved in the all-data solution, the solution
of our new velocities, having an eccentricity of
, was
compared with a circular-orbit solution. Although a close call,
according to the precepts of Lucy & Sweeney (1971), a circular
orbit is to be prefered. Thus, Table 5 lists the orbital
elements of UZ Lib, determined with the SB1C program and
our velocities only (one ESO velocity was given zero weight) and with
the period fixed from the all-data circular solution. Since the orbit
is circular, a time of maximum positive velocity is listed rather than
a time of periastron passage. Table A12 in the Appendix
lists our velocity observations and the residuals to the computed fit.
Shown in Fig. 9 is the computed orbit compared with our
velocities. The combination of very substantial rotational broadening
and line profile asymmetries results in a standard error of 3.0 kms-1
for an observation
of unit weight.
Copyright The European Southern Observatory (ESO)