The physical association of either a variable and a nonvariable star or two variable stars, is of great interest for the study of the origin and evolution of both stars. We are presently preparing a catalogue of such systems and have integrated some candidates in our current observational programmes.
One of these systems, IDS 23182S5421 (A = HD 220392/ B = HD 220391), is particularly interesting: from 41 measurements obtained in the seven colour photometric system of the Geneva Observatory, the standard deviation on the visual magnitudes of the brightest component, HD 220392 = HR 8895, appeared to be 0.011 mag. This is significantly larger than the observational noise level of 0.006 mag expected for a constant bright star (see: Table 4 in Rufener (1988)). Because its spectral type (F0IV) and the relevant photometric indices (d = 1.313, B2-V1 = 0.059) locate it in the d Scuti instability strip (Lampens1991), HD 220392 was reobserved for three weeks at the Swiss telescope at La Silla in autumn 1991. In the same sequence, HD 220391 was observed as well. Because of the large angular separation of the pair, both stars could easily be observed with a conventional photometer and there was no need to use a smaller diaphragm or to switch to a CCD detector.
Regular short-period light variations with a typical length of several hours (P ~ 5 hours) and a peak-to-peak amplitude of 0.05 mag in visual light have been detected in the preliminary data of HD 220392 (upper part of the figure). A preliminary periodogram analysis (PDM, Stellingwerf 1978) of about 130 data points, the majority of which was obtained during nine almost consecutive nights, indicates the presence of a period of about 0.214 days. The amplitudes of the fitted sine curve and the residual standard deviation are 0.024 and 0.010 mag, respectively. With such colours and periodicity, HD 220392 could well be a new d Scuti variable star. However, more data should be gathered and a periodogram-analysis of the definitive data should be awaited before reaching a conclusion about the type of variability.
The interest of observing this variable star furthermore lies in the fact that it very probably forms a physical pair with its wide companion star, HD 220391, at an angular separation of 26.2". From 14 micrometric measurements covering a period of 150 years - kindly provided by Dr. C.E. Worley from the Double Star Catalogues at U.S.N.O., Washington - it appears that this large separation is accompanied by a very small relative motion of the order of 0.0023", classifying the system as a common proper motion pair. The direction 122 of the relative motion doesn't fit well the difference of the proper motions of both components but the superiority in precision of the relative motion deduced from micrometric data over the proper motion data has been shown (Dommanget 1974).
Using the hereby adopted values (va = 0.0023" and sep = 26.2") and the known magnitudes and spectral types (comp. A: 6.15 - F0IVn; comp. B: 7.12 - A9V), one finds by using Dommanget's (1955) formula an inferior limit of 0.007" for the parallax of a possible physical system. Unfortunately, no trigonometric nor spectroscopic parallax is known that allows to distinguish between opticity or physicity. The Geneva photometric indices for HD 220392 permit to estimate the absolute magnitude through calibration (Hauck, 1973, 1985), giving Mv = +0.82 ˜ 0.43 with a corresponding photometric parallax of 0.009 ˜ 0.001". Therefore, the system should not be rejected as an optical pair.
By chance, both components are being observed by HIPPARCOS and very accurate parallaxes of both stars may soon be available. A better recognition of the physicity of this sytem will then be possible.
HD 220391 has colours very similar to those of HD 220392 (Gray and Garrison,1989), which also situates it in the region delimited by the d Scuti stars. From the first analysis, no obvious short period variability was found with an amplitude larger than 0.01 mag in the preliminary data of HD 220391 (see bottom part of the figure). Whether this companion is stable or variable at a smaller amplitude than the primary star must be investigated by accurate analysis of the definitive data.
Either way, since both components have very similar colours and almost identical rotational velocities (v sin i, A = 153 km/s; v sin i, B = 140 km/s) in addition to having a probable common origin and common proper motions, the system may be extremely important for a better understanding of the factors determine amplitudes and periodicities in d Scuti variable stars.
Assuming that HD 220392 is a d Scuti star, if HD 220391 show a very different pattern of variability, then very few factors are left that can be at the origin of a different "response" to variability. Also, since neither of these stars are known to have metallic lines, a detailed study of their difference in variability pattern may provide important clues about the presence of non-variable, non-metallic stars in the Delta Scuti instability strip.
Figure 1: Light curves of HD 220392 and HD 220391 on 24/25 sept 1994