Double stars
Present estimates of the percentage of double versus single stars point toward a ratio of 60% or higher. This is also confirmed by the high number of new detections coming from ground-based and satellite surveys. But double stars are seriously complicating both the observations and the elaboration of theorical models. It frequently happened and still happens that studies of double stars are being neglected in favour of studies of single stars, galactic and extra-galactic research. However double stars have a serious advantage: the orbits
These data are extremely useful to:
Stellar structure
Scientists try to understand what processes are going on in the interior of stars. Models of internal structure are therefore elaborated in which stellar masses play a very important role. The only possible direct observation is by observing double stars: indeed, from the knowlegde of the we are able to determine the total mass of the system. This information is an essential parameter on the way to modelling also the stellar evolution.
Stellar formation
By studying the distributions of the orbital elements such as the orbital period, the eccentricity, the mass ratio,... of double stars as a function of population or age, one obtains:
Diversity
The distributions of the orbital parameters show a very wide range in orbital period, (real) separation, mass ratio... There exists a panoply of widely differing objects with a variety of possible processes that may complicate the study of such systems. For example, in very close binaries interactions between the binary components may occur, such as: Such binaries can no longer be considered as two non-interfering stars at a common position in space: additional physical processes have to be taken into account. Additional observational constraints are obtained from their spectroscopic or light curve analyses.
Search for planetary systems ("exo-planets")
One can also look for systems with extremely low or high mass values. In the lower limit case, the studied systems may have a "brown dwarf" or a giant planet for companion. In the astrometric mode, similar to the detection of astrometric binaries, the reflex motion of the star gravitationally pulled by its planetary system is observed. This technique then provides a direct estimation of the mass of the planets. This search for "exo-planets" has been boosted a lot since the discovery of a solar-like star with a planetary companion (Mayor & Queloz, 1995 Nature 378, 355).


What is a double or binary star ?
It's a pair of stars in orbit around a common centre of gravity under their mutual gravitational attraction. The individual stars in the system are called components: component A is the primary or the brightest and component B is the secondary or the faintest (in general). There are several types of binaries according to the techique of detection used. Here we mention four of them.
Visual Double Stars
The two components can be resolved visually through a telescope. Such systems have mostly been investigated for astrometric purposes: highly accurate measurements of the position of the star on the celestial sphere are obtained in order to gain knowledge of the orbit characteristics as well as visual estimates of the difference of magnitude (measure of the brightness) between the two components. In more recent times, observers used photometric techniques to get high precision measurements of the global magnitude and colours of these systems.

Example: HIC A: 71683 B: 71681 - Name: alpha Centauri A and B - Ref: Belikov Catalogue, 1995
Astrometric Double Star
These are apparently single stars (not (yet) resolved). In this case the binarity is revealed by perturbations of the position with respect to the "background" or "reference stars". These systems are studied by photographic means: measurements show the motion of the photocentre (this is the weighted centre of the light intensity of the two components). From such a study we can obtain e.g. the luminosity, the semi-major axis of the astrometric orbit, the position of the center of mass of the system.

This figure shows the motion from 1987 to 1996 in right ascension and declination of HIC 87895 (In Hercules) as observed by the HIPPARCOS satellite.
Spectroscopic Double Stars
Systems where the components are too close to each other to be resolved as separate stars, are studied by spectroscopy (study of the stellar atmospheres, radial velocity measurements via the Doppler effect).The study of their spectra shows periodic shifts due to the orbital motion around a mean velocity which is the motion in the radial direction of the system as a whole. We thus can obtain the spectroscopic orbit. It gives us the orbital period, information on the orbit itself such as the eccentricity and very useful, the mass ratio of the system. For example, the left figure illustrates the radial velocity curves for alpha Aurigæ A and B.

Example: Star: HIC 24608(AB) - Name: alpha Aurigæ A and B - Ref: Pourbaix (private communication)
Eclipsing Double stars
When the orbit is aligned with respect to Earth, one component may occult the other one. These "eclipses" are seen as periodic perturbations (drops) on the light curve (luminosity as a function of time, e.g. HIC 31173 see figure). These double stars are called eclipsing binaries. From the study of their light curves, we have access to the period, the inclination, the luminosity ratio and the radii of the stars.

The Hipparcos Catalogue, Vol 12 (ESA SP1200), 1997. Only one of the two eclipses has been observed.
Example: HIC 31173(AB) Name:WW Aurigæ A and B Ref: Belikov Catalogue, 1995

THE CCD TECHNIQUE AND DOUBLE STARS
The CCD (Charge Coupled Device) is a detector based on the technique of photon counting. The CCD chip consists of light sensitive micro-cells called "pixel" (picture element). Each pixel transforms the received photons into charges (e-). The read out of the CCD is done by a process of charge transfer. The electric signal is then processed by an ADC to produce a digitized image (frame). The following picture shows you the image CCD of a cluster: it is M16 in the constellation Serpens.

Such an image is used by us to calibrate the scale of the CCD chip.
With a CCD, we can measure the angular separation between the two components of the double star, their positions and their magnitudes. We simultaneously obtain astrometric and photometric information. We thus study "intermediate" double systems, in other words double systems with angular separation between 1 and 10 arcsec (see table below).

Angular Separation
Technique<1"between 1" and 10">10"
Photoelectric-PhotometryGlobal informationGlobal informationinformation on the components
CCD photometryGlobal informationinformation on the componentsinformation on the components


In astrometry, we can acquire accurate relative positions between two components (with an accuracy of 5% of the pixel width) while in photometry we can measure very accurate differences of intensities between the components (with an accuracy of 1-2%) as well as total intensities.
Some Results

The digitized image is sullied by different effects: So we have to clean up all these imperfections. Additionally we subtract the sky value from the "target" (double star in our case). This is the first treatment of a raw CCD image. After that, a near-gaussian profile in 3 dimensions is fitted to the digitized image (Lorentz or Moffat profiles are used).
(Cuypers, Proceedings of the International workshop "Visual Double Stars: Formation, Dynamics and Evolutionary Tracks", ASSL Series, 223,1997)

From an astrophysical point of view, the most interesting fitted parameters are (see figure 4.3):

» the position (x,y) of the centre (with an accuracy better than pixel width/20)
» the Full Width Half Maximum (FWHM) (the "smearing" parameter called "seeing")
» the Intensity

These allow us to obtain accurate basic information on the double system e.g:. the relative positions, the magnitude difference . With a good photometric calibration one also has access to the absolute magnitudes of each component.