Search

The next generation of optical interferometers will provide high-resolution imaging of celestial objects by using either the aperture synthesis technique or the direct imaging principle. To determine the technical requirements, we have developed an interferometric test bench, called SIRIUS. To preserve the quality of the image, fast corrections of the optical path differences within a fraction of a wavelength have to be applied: this is the cophasing of the array, whereas making it coherent aims at stabilizing the optical path differences within a fraction of the coherence length. In the SIRIUS test bench, coherence and cophasing are achieved by fibred delay lines. Air delay lines are also used for the raw delay equalization. We present an original implementation of a piston sensor, called chromatic phase diversity, which is adaptable to any interferometer, whatever the configuration of the entrance pupil and the number of sub-pupils and whatever the interferometric combiner. Our method is based on the dispersed fringes principle and uses a derived version of the dispersed speckles method. The numerical simulation shows the performance of the method in terms of cophasing, accuracy and limiting magnitude. Experimental tests have been carried out both with optical turbulence and without. They show good results in both cases, despite some instrument-related limitations that can be eliminated. We show that our method is able to handle an amplitude of correction of ±11(λ/2) with an accuracy of ∼λ/30 over many minutes. [ABSTRACT FROM AUTHOR]

Active pre-main-sequence binaries with separations of around 10 stellar radii present a wealth of phenomena unobserved in common systems. The study of these objects is extended from classical T Tauri stars to the Herbig Ae star HD 104237. The primary has a mass 2.2 ± 0.2 M⊙ and secondary 1.4 ± 0.3 M⊙. Spectrointerferometry with the VLTI/AMBER in the K-band continuum and the Brγ line is presented.It is found that the K-band continuum squared visibilities are compatible with a circumbinary disc with a radius of ∼0.5 AU. However, a significant fraction (∼50 per cent) of the flux is unresolved and not fully accounted by the stellar photospheres. The stars probably do not hold circumstellar discs, in addition to the circumbinary disc, due to the combined effects of inner magnetospheric truncation and outer tidal truncation. This unresolved flux likely arises in compact structures inside the tidally disrupted circumbinary disc.Most (≳90 per cent) of the Brγ line emission is unresolved. The line-to-continuum spectroastrometry shifts in time, along the direction of the Lyα jet known to be driven by the system. The shift is anticorrelated with the Brγ equivalent width. It is shown that the unresolved Brγ emission cannot originate in the jet but instead is compatible with stellar emission from the orbiting binary components. The increase in the absolute value of the equivalent width of the line takes place at periastron passage; it could arise in an accretion burst, a flare or in the increase in effective size of the emission region by the interaction of the magnetospheres. The binary longitude of the ascending node is found to be Ω = (235 ± 3)° and the orbit retrograde.The origin of the jet is revisited. The tidal disruption of the circumstellar discs creates difficulties to ejection models that rely on stellar magnetosphere and disc coupling. A scenario of a stellar wind collimated by a circumbinary disc wind is suggested. [ABSTRACT FROM AUTHOR]