A small FORTRAN program for assessment of alignment tolerance in arbitrary sequences of flat mirrors

Mirrors are a common component in THz and sub-THz quasi-optical systems. The system designer always faces the need to specify alignment tolerances for manufacturing, but computing the impact of misalignments using commercial programs can be tedious, requiring lots of simulations. This program tries to alleviate the burden by performing a Montecarlo analysis for an arbitrary sequence of flat mirrors, specified with the positions of the central axis reflection points in a single reference frame. It is somewhat similar to [1]. The first point is the position of the source, the last an arbitrary point on the axis of the output beam. Source and mirrors are assumed to have random errors in orientation, with a uniform distribution function within the specified tolerance. All mirrors are assumed flat, which is of course a crude approximation, but for tolerance calculations of converging mirrors it usually provides a useful guidance. The error made with this approximation becomes smaller and smaller as tolerances are reduced, and results can be used as a starting point for more accurate simulations with commercial programs. The program also computes the polarization and its dependence on alignment tolerances. It is freeware and available in Fortran 90 source form at ISTP (former IFP) website alongside this report. Users are kindly requested to cite this report in their papers.