As the first prototype Carlina-1 already shows good results, we are thinking about a larger-scale project, wich will be a new step towards large imaging interferometers, and will provide original science.

Carlina-2 will reach 40 to 100 m baselines, dependig on the establishment site, and have about 100 sub-apertures. Helium balloon always appears to be a good solution to carry the gondola, but studies on others solutions like giant pylons or suitable natural relief like a valley are in course.

Scientific programs:

Although the Carlina interferometer must be understood as a protopype of a much more ambitious interferometric optical array, its 40-m baseline version will allow to reach an angular resolution up to 3 mas at visible wavelengths in a field of view of about 0.5 arcmin. These instrumental performances, although limited when compared to those of the giant interferometers now in operation like the VLTI or the Keck Interferometer, are nevertheless sufficient to obtain many new and original scientific results in the field of stellar astrophysics. Thus, accurate determination of some fundamental physical parameters bringing new constraints to stellar models can be reached reasonably.

-The components of close multiple stellar systems, known as spectroscopic binaries, can be resolved individually. Without the help of adaptive optic systems, the use of modern imaging numerical technics as speckle masking or triple correlation, getting rid of the atmospheric turbulence, could achieve bidimensional high angular resolution image reconstruction and resolve the stellar disc of each star and of the close circumstellar environment. Although double stars are complex objects both from the observational and the modelling point of view, the accurate determination of their orbits will provide fundamental data to refine the theory of single stars formation and evolution. In the case of interacting binaries, additional physical processes have to be taken into account due to the interaction between the two components. Mass transfer, tidal interaction, apsidal motion, mutual reflection are some main processes which can characterize such systems.

-The stellar discs of many evolved stars as red giants or supergiants will be resolved. Using high angular resolution imaging techniques combined with adaptive optics, some fine structures as granulation or bright spots will be detected at the stellar surface. These observations allow to determine the structure and the dynamics of the stellar atmosphere and can give some new insight especially for the study of the stellar convection or pulsation..

-For evolved hot or cool stars, the close circumstellar region will also be resolved. High angular resolution observations of the diffuse light or in some specific emission lines of the circumstellar envelopes will give unique information on the nature of the central source itself, often revealed as multiple, on the stellar mass loss and on the physical and chemical conditions surrounding the star (composition, creation and destruction of dust, evolution of winds...). Hydrodynamical simulations of mass flow winds foresee alterations of the symmetric geometry of the circumbinary envelopes. Bright small structures may appear in the flow pattern like hot spots or streams, spiral arms... Their observation will open a new domain of study, concerning the physical links between the formation and the evolution of the circumstellar envelope and of the central (and multiple) source.

First step: Speckle mode observations

Waiting for the implementation of an adaptive optics, speckle imaging in the visible and in the near-IR is possible with a Hyper telescope as with a conventional single aperture telescope.

Indeed, we have check the capabilities of a Hyper telescope in the speckle mode with end-to-end numerical simulations. Figure 1a and 1b shows a snapshot image of a binary star (separation= 10 mas, brightness ratio= 0.5) that a 40m-Carlina will provide at 1 micron, in the perfect case (no wave front errors) and with differential piston error due to the turbulence. The pupil densification factor is 20 (maximum is 32).

Thanks to speckle processing techniques as speckle interferometry (Labeyrie), speckle masking (Weigelt), etc., it is possible to recover the object. For instance, figure 1c shows the autocorrelation function of the binary star recovered by speckle interferometry from a sequence of 350 frames.

Mirror array used for the simulations

(Simulations O. Lardière)

Speckle interferometry is then an interesting observing mode for the first phase (without AO, nor coronagraph) of the project which can already produce scientific results on simple objects as binary stars, and even extra solar planets.

More complex and extended sources can be imaged in speckle mode by using the Weigelt’s speckle masking technique (triple correlation). Thanks to this powerful technique, stellar surface and envelope can properly be recovered. The granulation of a red super giant star could be seen with a 40m-Carlina in speckle mode, as well as close interacting binary stars.

Second step: Adaptive optic mode observations

The adaptive optic system will be implemented in a second step, allowing long exposure observations on faint sources. In the case of Carlina, as the primary segments size is about Fried parameter, only one actuator per segment is required. Indeed higher order residual errors do not affect the interference pattern. According to a new algorithm, developed by F. Martinache and V. Borkowski (Martinache F. 2004, Borkowski et al., 2004),  the piston sensing is made from a dispersed speckle image. This technique is particularly well suited for sparse apertures.

With this adaptive co phasing system, two outstanding scientific goals are reachable with Carlina:

The numerous apertures will give rich images with many pixels, allowing seeing details on resolved objects, as stellar surfaces and envelopes (Fig 2), close binary stars (Fig 3).

Fig 3:  Image of a close binary star (the main component is resolved).

(Simulations O. Lardière)

The gain provided by the adaptive optics is also interesting for exoplanets imaging, with a coronagraph removing the starlight, allowing reaching higher contrast than with speckle observations. We plan to characterise the pegasides discovered by radial velocimetry (51 Peg, Ups And, 47 Uma, 55 Cnc,…). We could obtain direct images of the whole planetary systems.

Some numerical simulations are under process to determine the exposure time required to detect smaller or more distant exoplanets as Jupiter-like or even telluric planets.

 _____________________________________

Main technical characteristics of Carlina-2:

One of the possibility for the establishment site is a natural concave site that we selected at Plateau de Calern, in the Cote d’Azur Observatory, in the south of France.

The primary mirror is constituted with several small spherical segments, dispersed along a virtual giant sphere, all directly anchored in the ground. This wide diluted aperture forms an interferometric image of the sky on the half-radius sphere. A focal optic is placed on this half-radius sphere to catch the image of the chosen star. A carbon gondola connected with a captive helium balloon carries this assembly. High elasticity module Kevlar cables links the balloon to the ground and to the gondola. Two ground-based winches pulling the gondola provide star tracking.

The focal optic is mainly constituted by a spherical aberration corrector, called Mertz corrector, and a pupil densifier. As the giant diluted primary mirror is spherical, the aberrant formed image is corrected by the Mertz corrector. The pupil densifier concentrates the energy in the central peak of the diffraction figure, otherwise this energy would be dispersed in many secondary peaks in the Fizeau mode recombination (Labeyrie 1996). A high-sensitive CCD camera is then placed at the densified focus. There are two ways to implant adaptive optic system: first one is to place adaptive mirror between the Mertz and the densifier, second one is to develop adaptive primary segments. This second solution, potentially more efficient, could be explored in collaboration with Arcetri Observatory (Italy).

Scheme of the Carlina-2 Hypertelescope prototype (40 m version)

Technical goals:

With the simplified prototype built at OHP, we have already demonstrated that we can easily obtain fringes with two close mirrors, and that the balloon stability is sufficient to keep the image on the camera without servo-loop. We also built and tested a Mertz corrector that gave satisfaction. Observations with a three elements diluted aperture are actually in progress, and the study of the pupil densifier is achieved.

At the end of the year 2005, we began the study and construction of a Carlina at real size. We will demonstrate that this concept, with no delay-lines, using ground stability, and having internal metrology for the active optic, is a decisive simplification for large imaging interferometers, with regard to classic solutions.