7 Summary and outlook

The purpose of this article is to present a powerful and versatile approach to dynamic modeling of the imaging process of an optical stellar interferometer operating in aperture synthesis mode. Therefore a set of numerical models has been developed and linked together to form a comprehensive analysis tool covering the various engineering disciplines of astronomical interferometry. Although most of the model components were specifically developed for the VLTI their modular and open structure allows their application to other related instruments. The modeling tool can be divided into two main components: (1) An optomechanical End-to-End model computes the time-dependent instrument response to an point source, i.e. the complex electric field distribution at the entrance of the interferometric beam combiner. This serves as input to (2) a Fourier optical model which simulates the imaging of extended objects. The final output of the complete model is the dynamic complex fringe visibility for a given baseline. The article describes the modeling approach illustrated by examples and some didactic simulation results obtained for the VLTI.

Future work intends a refinement of the optical modeling algorithms including support for deformable/segmented mirrors (adaptive optics) and further enhancement of the models for diffraction propagations. An atmospheric turbulence model will be included to supplement the existing disturbance models for atmospheric wavefront piston and tip/tilt, wind load and seismic noise. The various optical detectors and sensors have to be modeled in a realistic way taking sensitivity and noise aspects into account.

Eventually the program package can be useful during all preparatory and operational phases of a stellar interferometry project. In the planning phase it can be used for detailed demonstration of instrument feasibility. During the design and analysis phase the program enables a gradual refinement of the subsystem layouts (control, optics, structure) by studying their mutual interaction. Finally, in the operational phase it can be used for preparation of observations. A preceding simulation run can determine optimum values of instrument parameters (e.g. baseline configuration, integration time) for a given scientific objective.

Acknowledgements

We would like to thank Andreas Glindemann, Thilo Hannemann, and Pierre Kervella for fruitful discussions, and the anonymous referee for valuable comments. This work has been supported in part by the German Aerospace Center DLR (Dr. E. Bachem) under contract OISI FKZ 50 TT 9439.