8 HD81410 = IL Hya

8.1 Brief history

The light varibility of HD 81410, suspected by Cousins & Stoy (1963), was confirmed by Eggen (1973), who found a V amplitude of 0$.\!\!^{\rm m}$4 and a period of 25.4 days. From additional photometry Raveendran et al. (1982) discovered that a period of 12.86833 satisfied all available photometry and identified the star as an RS CVn binary. Combining 25 years of V-band photometry, Strassmeier et al. (1997) obtained a long-term average period of 12.7864 days.

Bidelman & MacConnell (1973) classified HD 81410 as K1 III and noted that it has strong CaII H and K emission and also that its Balmer lines are filled in. Radial velocity observations of Wayman & Jones (Eggen 1973) showed the star to be a single-lined binary, and Raveendran et al. (1982) found the velocities to be consistent with their photometric period. From spectroscopic and photometric evidence they suggested that the secondary is probably an F dwarf. Balona (1987) combined his numerous velocities with those of Collier Cameron (1987) and determined an orbital period of 12.908 days. Following the spectroscopic discovery of the secondary (Donati et al. 1997), Weber & Strassmeier (1998) obtained multiwavelength Doppler images of HD 81410 and computed the first double-lined spectroscopic orbit for it. Recently, using 85 published velocities, Raveendran & Mekkaden (1998) determined both circular and elliptical orbits for the primary.

8.2 Orbital elements

We have obtained 101 observations of HD 81410 from 1981 through 1998. Of that total, 31 observations were made by FCF at McDonald Observatory and KPNO and 80 by the Vienna group at NSO and KPNO. Some of the velocities have previously appeared in Fekel et al. (1986) and Weber & Strassmeier (1998). The values used in this paper are slightly different due to the assumption of different standard-star velocities. In addition the Vienna group's KPNO H$\alpha$ spectra of February-March 1995 have been remeasured.

Complicating the orbital analysis of this star is the primary's significant chromospheric activity, resulting in line profile variability, plus the fact that 10 different data sets, containing from two to 46 velocities and totaling 179 velocities, are available for the primary.

Preliminary orbital elements of the primary were computed initially for the FCF velocities with BISP and then refined with SB1. Similar independent orbital solutions were obtained for the velocities of Balona (1987), the Vienna group's NSO velocities, and the Vienna group's KPNO velocities obtained at 6430 Å, to determine the weights of the observations and any zero-point velocity shift for each data set.

The other six data sets contain only a very small number of velocities or do not have good phase coverage. Since the FCF velocities have the longest baseline, from 1981 to 1998, of any of the data sets, the orbital solution of those velocities provides an excellent "template'' for the other six data sets. Thus, the six sets of velocities were included with zero weight in a separate SB1 solution of the FCF velocities. The velocity residuals to the FCF-velocity solution for each of the six data sets were used to determine the appropriate weights and zero-point shifts. The results for all 10 data sets are summarized in Table 4, where Col. 1 identifies the data set, Col. 2 notes the years covered, and Col. 3 gives the number of velocities. Columns 4 and 5 list the weight assigned to the velocities in each set and each set's zero-point shift, respectively.

 

Table 4: HD 81410 primary data sets 

An SB1 orbital solution of all the appropriately weighted velocities of the primary resulted in an eccentricity of $0.0046\pm 0.0027$. Because of this very low eccentricity, a circular-orbit solution for the same data was computed with SB1C. According to the precepts of Lucy & Sweeney (1971), the circular orbit is to be preferred.

Very weak lines of the secondary are visible in 35 of our spectra that have the highest signal-to-noise ratios. Independent SB1 solutions, assuming the elements of the primary except for the center-of-mass velocity and semi-amplitude, were obtained for the FCF velocities and the Vienna group's KPNO velocities. Those 35 secondary velocities plus two of Donati et al. (1997) were given a weight of 0.1 and no zero-point correction was applied.

An all-data circular orbit for the primary and secondary velocities together was computed with SB2C. Six velocities of the primary and one of the secondary were given zero weight because of their large residuals. The period from that solution was assumed, and a final SB2C solution was obtained with all velocities having weights $\geq$0.1. Those orbital elements are listed in Table 5. Since the eccentricity is zero, the time of periastron is undefined and has been replaced with the time of maximum positive velocity.

 

Table 5: Orbital elements 

Table A7 in the Appendix lists our velocities of the primary, those of Donati et al. (1997) as well as those obtained by Dadonas (1994), at Maidanak, Uzbekistan (Dadonas et al. 1994). Table A8 in the Appendix gives our velocities of the secondary and the two values of Donati et al. (1997). Figure 6 shows the computed velocity curves compared to the velocities listed in our two tables. The standard error of an observation of unit weight for the primary is 0.7 kms^-1, quite good for such an active star.

 

Figure 6: Radial-velocity curves for HD 81410. Filled circles are from KPNO, open circles from NSO, pluses from McDonald. Additionally, we added ten velocities from Dadonas (1994) (crosses) and two velocities of both the primary and secondary from Donati et al. (1997) (triangles). Zero phase is the time of maximum positive velocity of the primary