Eclipsing Binary Update Number 11

Number 11 April 2001

Notes from the Chairman

The most recent compilation of minima, Observed Minima Timings of Eclipsing Binaries No. 6, was published in October. This publication, coming to fruition largely through the efforts of Gerry Samolyk, lists 966 new minima of 50 stars and displays their O-C curves. The O-C curves for some of these stars now span 35 years and nicely illustrate the period changes these stars have undergone during that time.

Several amateur CCD observers, taking their work quite seriously, have honed their technical skills and are obtaining high quality light curves useful for modeling some of the recently discovered eclipsing binary stars. An ad hoc AAVSO observing team consisting of visual observers, amateur CCD observers and professionals acting as both observers and advisors have tackled a number of eclipsing binaries recently discovered via the Robotic Optical Transient Search Experiment (ROTSE). Typically the visual observers find when eclipses occur and establish a rough ephemeris, the amateur CCD observers refine the period and establish high quality light curves, the professionals provide technical advice and obtain color photometry for the variable and comparison stars; and a trip to the Harvard Plate Stacks results in finding archival minima that extend the baseline to further refine the star's period. You can find out more about the ROTSE variable star discoveries on the Web at:

http://www.umich.edu/~rotse/

The Eclipsing Binary Ephemeris is a feature recently added to the AAVSO Web site. Kate Davis, technical assistant at HQ, scanned the ephemerides for 2001 and placed them on the AAVSO Web site in Adobe pdf format. You can find it at:

http://www.aavso.org/committees/eb/ebephem.stm

We expect that all the above subjects plus a review of procedures for visual observing and processing of data will be discussed in some detail at the upcoming Eclipsing Binary Workshop, which will take place on 4 May 2001 at the Spring Meeting of the AAVSO at Madison, Wisconsin. Mark it on your calendar, plan to be there, and bring your questions and comments. We look forward to your participation. -- MEB

Two's Company, Three's a Crowd - Part 3

In the final installment of this extended essay (the first two parts appeared in EBU Nos. 8 and 9), I want to mention some further phenomena that can occur in triple systems with eclipsing binaries and that sometimes confuse and sometimes enlighten astronomers who study them. All of these phenomena are related to apsidal rotation. The "line of apsides" connects the periastron and apastron of an elliptical orbit (i.e., the major axis of the ellipse). Many detached eclipsing binaries have elliptical orbits and as a result (except for the special case when the line of apsides points straight at us), the secondary minimum is displaced from phase 0.5. This asymmetric position of the secondary minimum is the telltale sign of an eccentric orbit.

There are wheels within wheels, of course, and rotation of the line of apsides has been detected in dozens of eclipsing binaries. In systems with apsidal rotation, the primary and secondary minima oscillate about their mean values in the O-C diagram. Astronomers are interested in apsidal rotation because it offers clues to the internal structure of the orbiting stars. The period of apsidal rotation can extend from a few decades to many centuries. The rotation rate is driven by contributions from both classical and the relativistic mechanics.

When the binary is joined by a third stellar component, the situation gets complicated. The observed apsidal rotation rate will differ from the theoretical rate due to the light time effect as the binary orbits the center of gravity of the triple system. M. Wolf, R. Diethelm, and L. Sarounova (Astron. Astrophys. 345, 553) recently published light-time analyses for two well-known eclipsing binaries with apsidal rotation, RU Mon and DR Vul. Taking account of such factors as the inclination and eccentricity of the third-body orbit, the calculations required the determination of 10 independent variables. The two figures below show the O-C residuals (left) of the primary and secondary minima of RU Mon with respect to linear light elements and the light-time effect (right) after subtracting the apsidal motion. The third stellar component has an orbital period of 73 years with an eccentricity of 0.46.

Click image to enlarge.

In some eccentric binary systems, the relativistic contribution to apsidal rotation can exceed that of classical mechanics. About 20 years ago, considerable interest was stimulated when it was noticed that the observed apsidal rotation rate of some eclipsing binaries, such as DI Her and AS Cam, differed significantly from the theoretical rate. Classical gravitational theory was well established, but could the theory of relativity be wrong in some part? Published data was collected and many new observations were made. In some cases, the addition of new, high-precision data appears to have resolved the problem. For example, I. M. Volkov and Kh. F. Khaliullin (IBVS 4680) found that the apsidal rotation rate of V541 Cyg is not significantly slower than theory, as was previously thought.

In other cases, new data haven't just improved the observed apsidal rotation rate to match theory but have revealed a new explanation for the discrepancy between the observed and theoretical rates - the gravitational influence of a third body. In the case of AS Cam, for example, V. S. Kozyreva and Kh. F. Khaliullin (Astron. Reports 43, 679) have used many new, precise times of minima to confirm that the observed apsidal rotation rate is only one third of the theoretical value. But their analysis of the times of minima also revealed synchronous cyclic variations in the times of primary and secondary minima. These variations can be explained by a third body in an eccentric (e = 0.5) orbit with a period of 805 days, which changes the gravitational dynamics of the whole system. Further observations and analysis will probably show that the same explanation applies to any remaining cases where the observational and theoretical rates differ.

In recent years, a growing number of eclipsing binaries have been found to be members of triple and higher-order systems. Indeed, it almost seems that any eclipsing binary that is investigated closely enough will ultimately be found to have a third component. Whatever the truth may be, the recognition of many eclipsing binaries in triple star systems has increased the potential value of careful timings of minima and has expanded the range of information that can be gleaned from the study of eclipsing binaries. --DBW

Amazing: the visual discovery of a naked-eye eclipsing binary

On July 1, 1997, Sebastian Otero of Buenos Aires noticed that delta Vel was about 0.3 magnitude fainter than usual. He began monitoring the star and detected three more minima in 1998 and 1999. He was trying to work out the period when he learned in October 2000 that Paul Fieseler at Jet Propulsion Laboratory had noted a dimming of delta Vel in the star scanner data of the Galileo satellite. A look back through the Galileo archives uncovered another recorded minimum in 1989. An alert notice was issued on vsnet, and observers in Australia and South Africa helped to find more minima.

Determining the period proved to be difficult due to the binary's eccentric orbit. But in IBVS 4999, Otero, Fieseler, and Lloyd were able to report that delta Vel's period is 45.15 days with the secondary minimum occurring at phase 0.43. Eclipses are about 1 day long, with primary minimum 0.45 V and secondary minimum 0.26 V deep. The discovery of delta Vel's variability is probably the first visual detection of a naked-eye eclipsing binary since J. F. J. Schmidt discovered u (68) Her in 1869-and that star is 5th magnitude, not 2nd!

VW Pegasi defeated at last!

M. Dahm discussed all this confusion in the BAV Rundbrief in 1996, and three other BAV observers, H. Achterberg, P. Frank, and D. Husar, decided to settle the question once and for all. During three seasons beginning in 1997, these observers obtained more than 5,000 CCD measures and caught 11 minima. Their thorough analysis showed that Kämper was right, the period is 21.07 days. Thanks to the precision of CCD observations, the BAV team also discovered that the observed minima were not equal in depth or duration, and this allowed them to distinguish between the primary and secondary eclipses (light curve, below). They found that the primary minimum has a depth of 0.71 magnitude and the secondary 0.65 magnitude in red (unfiltered CCD) light. The secondary eclipse is also about 40 percent longer than the primary eclipse and occurs at phase 0.27, indicating an orbital eccentricity of e = 0.39. The complete report appeared in IBVS 4916.

The Strange Case of Theta-1 Orionis A

D. J. Strickland and C. Lloyd reviewed the status of Theta-1 Ori A (also now V1016 Ori) in The Observatory, April 2000, p. 141. They extracted radial velocities from the International Ultraviolet Explorer satellite archive to calculate a new orbital solution and analyzed times of primary minimum for new light elements.

At infrared wavelengths, Theta-1 Ori A has a very close (0.2 arcsec) companion. The system also includes a compact and variable radio source of puzzling origin. The orbit is highly eccentric (e = 0.63). To date, the secondary minimum has not been detected at any wavelength, and this remains the single largest hindrance to defining the system parameters. Radial velocity and partial light curve solutions indicate that the primary star has the temperature, radius, and mass of a B1V star. But a secondary star with a range of cooler temperatures can fit the light curve of primary minimum equally well.

Assuming an appropriate mass of 12 solar for the primary star, the radial velocities indicate that the secondary mass is 3 solar. This corresponds to a spectral type near B9, but the derived radius is too large for this mass and suggests that the secondary star is not a main sequence object. Since the age of stars in this region is only 1-2 million years, the secondary star is not an evolved star expanding away from the main sequence as in ordinary Algol-type systems but a young star that is still contracting toward the zero-age main sequence (ZAMS).

Click image to enlarge.

The Even Stranger Case of LD 345

AAVSO observer Pete Guilbault took an interest in the star and organized an international effort to define its variability. Pete and Tim Hager checked the Harvard patrol plates, and N. E. Kurochkin examined the Moscow plate collection, but neither observatory had plates taken during the 1972-74 interval and no other minima were found between 1905 and 1995 (with gaps from 1908-28, 1952-62 and 1962-75). The search was then widened to Asiago Observatory, where D. Moro measured 6 Schmidt plates exposed between 1967-70 and 10 plates between 1974-76. At Sonneberg, E. Splittgerber and P. Kroll checked 111 photovisual plates taken between 1970-77.

The Sonneberg plates showed LD 345 at maximum except from December 1971 to September 1974, when it was invisible, fainter than 13.0 pv. The Asiago plates showed the variable at maximum in 1970 but at 16.8 ptg in July 1974, brightening to 15.6 ptg in September and 14.7 ptg in November. LD 345 was back at maximum in May 1975.

From all this material, the investigators found that the minimum from 1971-74 lasted 1000-1200 days, and the recovery to maximum lasted no more than 150-200 days. Assuming a symmetrical light curve with a constant interval at minimum, the constant phase lasted 700-800 days. If LD 345 is an eclipsing variable and the minima occur at equal intervals, the period must be at least 47.1 years, twice the period of epsilon Aurigae. If minima occurred during the gaps in the photographic record, then the interval between minima must be irregular.

Arne Henden at the Naval Observatory's Flagstaff station measured the colors of LD 345 at maximum and found color indices typical for an M0III star but atypical for R CrB or symbiotic variables. The photographic observations of one long, deep minimum cannot prove that LD 345 is an eclipsing binary, but the photometric behavior and color do not make a strong case for other types of variables with eclipse-like minima. Spectroscopy may solve the puzzle, but until then, classification of LD 345 remains uncertain. In the meantime, visual observers are monitoring this odd variable. For information, contact Pete Guilbault (pete1199@aol.com). The results of this investigation appeared in IBVS 4926.

Eclipsing Binary News & Notes

The Team's newest publication is IBVS 4998, in which they present improved light elements and a preliminary light curve solution for the 0.51-day eclipsing binary GSC 1534-0753. This star is another discovery of the ROTSE1 CCD survey. Marv Baldwin worked out preliminary light elements from visual monitoring, Pete Guilbault and Tim Hager gleaned many additional times of minima from the Harvard patrol plates, and CCD observers Gil Lubcke and Gary Billings contributed a complete V light curve. Arne Henden at the Naval Observatory Flagstaff Station determined UBVRI magnitudes of the comparison star and of the variable at several phases, and Dirk Terrell at the Southwest Research Institute ran the observations through the Wilson-Devinney light curve analysis program. The solution indicates that GSC 1534-0753 is in marginal contact with the less-massive star eclipsed at primary minimum. The poor thermal contact suggests that the binary is a B-type W UMa system. The Team is now acquiring a multi-filter CCD light curve for DU Leo and has several more stars to work on this summer.

The eclipsing binary EI Cep is another Bamberg discovery. The system has almost equal primary and secondary minima, range 7.5-8.0 V, an F1 spectral type, and a period of 8.4 days. As a result of a new study by G. Torres, J.Andersen, B. Norström, and D. Latham (Astron. J. 119, 1942), EI Cep has joined the small but growing list of binary systems with precisely determined masses and radii. This binary system is of particular interest because both stars are significantly evolved from the zero-age main sequence and are close to the end of the core hydrogen-burning phase. Thanks to high-precision radial velocity and light curve solutions, the masses are now known to better than 0.4 percent and the radii to 2 percent. The primary (more massive) star is the larger and cooler member of the pair. The secondary star is particularly interesting because it displays the chemical anomalies of metallic-line A and F stars. The masses of the two stars differ by just 5 percent, the effective temperatures differ by only 200 K, and the rotation rates of both stars appear to be nearly synchronized with the orbital motion, yet one star shows the Am anomaly and the other does not.

V1073 Cyg is yet another bright (8.2-8.6 V) Bamberg discovery, an A-type W UMa system with a period of 0.79 day. Several light curve solutions have been published, but they were based on a faulty spectral classification of A3V for the primary component. But a re-examination of the spectrum and multi-filter photometry indicates that it should be classed as early F. In a new study of the system, S. Morris and S. Naftilan (PASP 112, 852) find that the mass and radius of the primary are appropriate for a late-A or early-F star that has evolved beyond the terminal-age main sequence. This is supported by a period decrease of about 0.8 second in 1982. Assuming conservative mass transfer within the system, the decrease indicates that mass was transferred from the evolving primary to the smaller secondary component.

Featured star: V1010 Ophiuchi

The primary component is spectral type A7IV-V, and G3V has been inferred for the secondary star. There is a 40-minute interval of nearly constant light at minimum as the smaller, cooler secondary transits the disk of the larger, hotter primary (light curve, previous page). Various authors have described V1010 Oph as an "early-type contact binary," "short-period, EB-type binary," "A-type W UMa" and "EB-type W UMa" variable. J. S. Shaw nominated V1010 Oph as the prototype for a small but distinct subclass of thermally decoupled, near-contact binary systems near the beginning of their evolution toward contact status.

Your Editor glories in the distinction of being the first person to observe a minimum of V1010 Oph in real time, using 7x50 binoculars during the summer of 1965. The then-editor of Sky & Telescope, Joseph Ashbrook, thought that the announced period might be in error and suggested that he check it. The minimum occurred on time, and the period proved to be correct. But there are degrees of correctness. When Your Editor observed V1010 Oph again in 1982, he found that the minima were occurring almost three hours earlier than predicted. The period announced at discovery had been determined by assuming a constant period between two small sets of minima separated by 30 years. When the O-C curve was filled in, however, it exhibited the strong negative curvature of an inverted parabola. The period of V1010 Oph has been decreasing steadily by 0.03 second per year for at least the past 80 years. This may not seem like much, but it is enough to shift the time of mid-eclipse by more than an hour per decade.