Radial velocities for AG Dor were determined from cross-correlations
with spectra of the radial velocity standard star HR1829
(
kms
; Blumberg & Boksenberg [2000]),
observed during the same night. The cross correlations were computed using
IRAF's fxcor routine (see Fitzpatrick [1993]), which fits a
Gaussian to the cross-correlation function to determine the central shift.
Because the H
line can be severely
affected by velocity fields originating from microflares or a stellar wind,
the H
region was excluded from the cross correlations.
However, the secondary of AG Dor shows up only as a weak emission line in
H
. Its radial velocities were measured from the residual spectra
after subtraction of a reference spectrum (Sect. 3.2).
In case of UX For, FeI
6546 Å is the only line that
significantly contributes to the cross-correlation function
besides the H
line. Consequently, the radial velocity
of
6546 was measured directly by fitting a double Gaussian
to the primary and secondary line using IRAF's splot
routine. In seven out of 37 cases, the primary and secondary
lines were severely blended and no reliable
radial velocities could be obtained. All other
observations were given equal weight when calculating the orbit.
Orbital elements were derived with the differential-correction program
of Barker et al. ([1967]), as modified and described by Fekel et al.
([1999]). Our final elements for both binaries converged at an
eccentricity so close to zero that a formal zero-eccentricity solution was
adopted (see Lucy & Sweeney [1971]).
The standard error of an observation of unit weight is 4.62 kms
for UX For, using all available measurements of the primary and
secondary component. Some of Balona's ([1987]) O
C residuals are
12 kms
, and two of Balona's secondary O
C residuals are as
large as 23 kms
. The standard error for the AG Dor primary was 0.41 kms
,
and was calculated using all 38 cross-correlation measurements.
The secondary-star orbit was computed independently from the primary star
by using the radial velocities of the residual H
emission feature
(see Sect. 4). H
is the only line that is detected from the secondary
star and the accuracy of the secondary measurements is comparably
poor. Its standard error of an observation of unit weight is 8.6 kms
.
Our revised elements agree very well with the preliminary orbit (from L.
A. Balona [1987]; in Strassmeier et al. [1993]). Our best
orbits with minimized O
C residuals were adopted as the final solution,
are given in Table 3 and plotted in Fig. 2
and Fig. 3 for UX For and AG Dor, respectively.
Phases of all line profiles were computed using revised ephemerises. For UX For
![]() |
![]() |
| UX Fornacis | AG Doradus | ||||||
| Element | Prelim. | Final | Error | Prelim. | Final | Error | Unit |
|
|
0.95479 | 0.9548120 | 2.562 | 2.5620 | 0.0027 | days | |
| 19.1 | 20.3 | 0.7 | 70.2 | 68.95 | 0.08 | kms |
|
| 91.8 | 95.6 | 1.2 | 58.0 | 56.13 | 0.13 | kms |
|
| - | 131.0 | 2.8 | - | 97.9 | 1.6 | kms |
|
| 0 | 0 | - | 0.04 | 0 | - | ||
| 4000.512 | 8945.2348 | 0.002 | 4378.67 | 8945.496 | 0.002 | 244+HJD | |
| 0.077 | 0.087 | 0.003 | 0.052 | 0.0470 | 0.0003 | M |
|
| - | 1.371 | 0.024 | - | 1.745 | 0.032 | ||
| - | 0.667 | 0.023 | - | 0.617 | 0.024 | M |
|
| - | 0.487 | 0.014 | - | 0.354 | 0.009 | M |
|
| 1.21 | 1.255 | 0.015 | 2.04 | 1.977 | 0.005 | ||
| 1.48 | 1.721 | 0.037 | - | 3.45 | 0.06 | ||
| Standard error of an | |||||||
| observation of unit weight | 4.62 | 0.41/8.6 | kms |
||||